University of South FloridaScholar Commons

Graduate Theses and Dissertations Graduate School

January 2012

Investigation into the Mechanism of Salicylate-Associated Genotypic Antibiotic Resistance inStaphylococcus aureusNada Salah HelalUniversity of South Florida, nhelal@mail.usf.edu

Follow this and additional works at: http://scholarcommons.usf.edu/etd

Part of the American Studies Commons, Microbiology Commons, and the Molecular BiologyCommons

This Thesis is brought to you for free and open access by the Graduate School at Scholar Commons. It has been accepted for inclusion in GraduateTheses and Dissertations by an authorized administrator of Scholar Commons. For more information, please contact scholarcommons@usf.edu.

Scholar Commons CitationHelal, Nada Salah, "Investigation into the Mechanism of Salicylate-Associated Genotypic Antibiotic Resistance in Staphylococcusaureus" (2012). Graduate Theses and Dissertations.http://scholarcommons.usf.edu/etd/4069

 

Investigation into the Mechanism of Salicylate-Associated Genotypic Antibiotic

Resistance in Staphylococcus aureus

by

Nada S. Helal

A thesis submitted in partial fulfillment

of the requirements for the degree of Master of Science

Department of Cell Biology, Microbiology and Molecular Biology College of Arts and Sciences University of South Florida

Major Professor: James Riordan, Ph. D. Kathleen Scott, Ph. D. MyLien Dao, Ph. D.

Date of Approval: June 6th, 2012

Keywords: Microbiology, NSAIDs, Bacteriology, Fluoroquinolones, Gram-positive

Copyright © 2012, Nada S. Helal

     

DEDICATION

I dedicate this to my husband, James, for being there for me throughout this degree.

Your patience and belief in me are the reasons behind my success. I would not have been

able to complete this degree with out you. I would also like to dedicate this to my parents

for their faith in me, my sister, Noha, for her tough love and my friends, Iyat, Mariana and

Niti for their moral support and for allowing me to vent over the years. Thank you all for

helping me complete my degree. A special dedication to my former undergraduate, Jessica

Cheer, who was a great help in this study and will truly be missed.

     

ACKNOWLEGEMENTS

I would like to acknowledge my advisor, Dr. James Riordan, for his support,

guidance and patience with me during the past three years. Your expertise has helped me

become a knowledgeable micro and molecular biologist. I would like to acknowledge Dr.

Lindsey Shaw for his assistance in understanding Staphylococcus aureus as well as his

generous contribution of strains used in this study. I would like to acknowledge Dr. Lucas

Li and his lab for their aid in the metabolomics project. I would like to acknowledge Dr.

John Gustafson for his generous contribution of strains used in this study as well as the

collaborative work in our recent publication. I would like to acknowledge my committee

members for their guidance and time during my degree. I would like to acknowledge the

Riordan lab for their continuous guidance and aid in protocols.

   i  

TABLE OF CONTENTS

List of Tables .......................................................................................................................... iii List of Figures ......................................................................................................................... iv Abstract ................................................................................................................................... v Chapter One: Introduction ..................................................................................................... 1 NSAIDs and salicylate ............................................................................................... 1 The effects of salicylate on eukaryotes ...................................................................... 1 The effects of salicylate on bacteria .......................................................................... 2 Staphylococcus aureus ............................................................................................... 4 Antibiotic Resistance in S. aureus ............................................................................. 5 Salicylate associated phenotypic and genotypic antibiotic resistance in S.

aureus .................................................................................................................. 6 Hypothesis and aims of the study .............................................................................. 8 Chapter Two: Characterization of the Salicylate-Associated Genotypic Antibiotic

Resistance Phenotype in S. aureus ...................................................................... 10 Background ................................................................................................................ 10 Methods ..................................................................................................................... 12 Bacterial strains and culture media and conditions ....................................... 12 Antibiotics and NSAIDs ................................................................................ 12 Determination of mutation frequency ............................................................ 13 Antibiotic susceptibility by the minimum inhibitory concentration

assay ......................................................................................................... 14 S. aureus chromosomal DNA extraction ....................................................... 14 Sequencing of target site modifications in antibiotic resistant

mutants ..................................................................................................... 15 Selection for resistance to sodium salicylate ................................................. 16 Spontaneous selection .............................................................................. 16 N-methyl-N-nitro-N-nitrosoguanidine-mutagenesis ............................... 17 Stepwise selection .................................................................................... 17 Results ....................................................................................................................... 18 Antibiotic specificity of the salicylate-associated genotypic

resistance phenotype ................................................................................ 18 Antibiotic specificity of the SAGAR phenotype ........................................... 21 Characterization of mutations conferring resistance to

fluoroquinolones ...................................................................................... 23

   ii  

Role of sigB and mgrA in the SAGAR phenotype ........................................ 25 Investigating the chemical signature associated with salicylate-

associated genotypic resistance to antibiotics .......................................... 26 Discussion ................................................................................................................. 28 Chapter Three: Role for metabolic stress in the Salicylate-Associated Antibiotic Resistance Phenotype ............................................................................................................ 34 Background ................................................................................................................ 34 Methods ..................................................................................................................... 36 Generation time determination ...................................................................... 36 Mutation frequency determination ................................................................. 36 Metabolite profiling ....................................................................................... 37 Analysis of intracellular metabolites ................................................. 38 Reactive oxygen species assay ...................................................................... 38 NAD+/NADH assay ....................................................................................... 40 Results ....................................................................................................................... 41 Dose-dependence of salicylate-associated genotypic antibiotic

resistance .................................................................................................. 41 Metabolite profile of S. aureus in the presence of salicylate ......................... 44 The effect of salicylate on the TCA cycle ..................................................... 51 Oxygen dependence on SAGAR phenotype .................................................. 52 Role for reactive oxygen in the SAGAR phenotype ...................................... 53 Discussion ................................................................................................................. 57 References .............................................................................................................................. 62

   iii  

LIST OF TABLES

Table 2.1: Bacterial strains and plasmids .............................................................. 13 Table 2.2: Primers used in this study ..................................................................... 13 Table 2.3: Antibiotic specificity of salicylate-associated genotypic phenotype .... 20 Table 2.4: Minimum inhibitory concentrations (MICs) for antibiotic resistant

isolates selected with or with out salicylate ......................................... 23 Table 2.5: Sequencing results for grlA in fluoroquinolone resistant isolates in

SH1000 compared to NCBI S. aureus sp. N315 ................................... 24 Table 3.1: Metabolite profile for S. aureus grown with salicylate ........................ 47

   iv  

LIST OF FIGURES

Figure 2.1: Frequency of resistance to ciprofloxacin in S. aureus cultures adapted to salicylate ........................................................................... 21 Figure 2.2: Structural differences between salicylate, benzoate and acetyl salicylic acid ...................................................................................... 27 Figure 2.3: Dependence on salicylate chemical structure for SAGAR ................. 28 Figure 3.1: Dose dependency of salicylate, aspirin, and benzoate for increased frequency of resistance to ciprofloxacin ............................ 43 Figure 3.2: Effect of salicylate on cellular NAD+ levels ....................................... 52 Figure 3.3: Expression of SAGAR during anaerobic growth ................................ 53 Figure 3.4: Salicylate associated ROS accumulation ............................................ 56 Figure 3.5: The effects of glutathione on the SAGAR phenotype ......................... 57

   v  

ABSTRACT

Growth of Staphylococcus aureus with the NSAID salicylate increases phenotypic

resistance (SAPAR), and the frequency at which heritable resistance occurs to various

antibiotics (SAGAR). This study describes the effect of salicylate on heritable and

phenotypic resistance to a set of antibiotics for laboratory and multi-drug resistant strains

of S. aureus and investigates the link between resistance and SAGAR. Drug gradient plates

were used to determine phenotypic resistance to antibiotics targeting DNA replication,

transcription, translation and the cell wall in the presence or absence of salicylate. To

measure heritable resistance, mutation frequencies were determined for each antibiotic in

the presence and absence of salicylate. Salicylate significantly increased mutation

frequency of SH1000 to ciprofloxacin 27- fold from 4.9 x 10-8 to 8.5 x 10-7. A significant

8.5- fold increase was observed for LAC from 5.2 x 10-7 to 2.1 x 10-6. Conversely,

salicylate significantly decreased mutation frequency for SH1000 to lincomycin 0.035-fold

from 3.4 x 10-7 to 1.3 x 10-7. Deletion of the general stress sigma factor sigB encoding σB

in SH1000 resulted in decreased heritable and phenotypic resistance, signifying the

importance of σB in the full expression of both phenotypes. Metabolite profiling revealed

downregulation of glycolysis, TCA, pentose phosphate pathway, and amino acid

metabolism. The downregulation of the TCA cycle was confirmed as observed through an

increase in NAD+ at growth toxic concentrations of salicylate. Salicylate has been shown to

result in ROS accumulation and disruption of proton motive force in mitochondria.

   vi  

SAGAR was only detected for fluoroquinolones, which have been shown to impair TCA

cycle and result in ROS accumulation. Examination of ROS under growth-toxic

concentrations of salicylate did not reveal a significant increase in ROS levels. Also, the

combination of ciprofloxacin and salicylate did not result in an increase in ROS levels.

Despite this, addition of the antioxidant glutathione abrogated SAGAR for ciprofloxacin in

SH1000 but not for SAPAR. Analysis of SAGAR with NSAIDs benzoate and acetyl

salicylic acid revealed a necessity for the ortho hydroxyl group on salicylate to fully

express SAGAR. These results suggest that salicylate has pleiotropic effects on S. aureus

that include antimicrobial resistance, altered metabolic flux and accumulation of ROS as

well as unidentified regulatory genes.

   1  

1  

CHAPTER ONE: INTRODUCTION

1.1 NSAIDs and salicylate

Non-steroidal anti-inflammatory drugs (NSAIDs) have analgesic and anti-

inflammatory properties and are used in the treatment and control of inflammatory

conditions and in the treatment of various cancers [1-5]. NSAIDs inhibit

cyclooxygenases, a class of enzymes responsible for producing mediators of

inflammation including prostaglandins, prostanoids, thromboxanes. NSAIDs are among

the most widely prescribed drugs, with approximately 100 million prescriptions filled

annually in the USA alone [6]. Acetylsalicylic acid (aspirin) was the first synthesized

NSAID, and is known for its extensive analgesic, anti-pyretic and anti-inflammatory

activity [2, 3]. Salicylate, the primary metabolite of aspirin, is also recognized for its

strong analgesic and antipyretic properties [3, 7] and in addition, is broadly used in oral

and topical medications, as a preservative in foods, and in various commercial

applications [8].

1.2 The effects of salicylate on eukaryotes

Salicylate and its chemical derivatives have been used since 400 B.C for their

analgesic and antipyretic properties, and more recently as antiplatelet agents for the

prevention of myocardial infarction and stroke [9, 10]. Salicylate and acetyl salicylic acid

(aspirin) have also been shown to prevent colon cancer [11], and to have chemoprotective

properties against lung and breast cancer [12]. This is believed to result from

   2  

enhancement by salicylate of the mitochondrial permeability transition-dependent

apoptosis (MPT), which acts by promoting apoptosis of transformed cells [12].

Salicylate has been shown to decrease the threshold for the onset of mitochondrial

permeability transition (MPT) [13, 14]. The MPT involves the formation of a non-

specific pore across the inner membrane permitting the free distribution of ions, solutes,

and small-molecular-weight molecules across the membrane [15]. The collapse of the

mitochondrial membrane potential and uncoupling of the electron transport chain from

ATP production has been shown to promote MPT. This disruption or collapse is also

associated with the loss of matrix calcium and glutathione, increased oxidation of thiols,

and further depolarization of the inner mitochondrial membrane, which increases the

gating potential for the MPT pore [15-19]. Salicylate can induce MPT at low

concentrations, resulting in an increase in the vulnerability of rat hepatocytes to necrosis

from oxidant stress [11], while high concentrations of salicylate (>3 mM) can lead to

MPT and cell death. It is believed that salicylate-dependent onset of the MPT may be

responsible for Reye’s syndrome [11]. Reye’s syndrome is a rare and severe illness in

children with a mean mortality rate of 40% [20]. The etiological cause is presently

unknown. However, this disease is believed to correlate with the use of salicylates and/or

other anti-pyretic drugs [21-24].

1.3 The effects of salicylate on bacteria

Salicylate has been shown to induce a number of distinct morphological and

physiological changes in bacteria. Importantly, when grown in the presence of sub-

inhibitory concentrations of salicylate, clinically significant species of bacteria including

   3  

Escherichia coli, Staphylococcus aureus, Pseudomonas aeruginosa and others, express

increased levels of intrinsic antibiotic resistance [7, 25-27]. Importantly, these effects are

induced under concentrations of salicylate that do not impair bacterial growth rates,

suggesting that they are specific to salicylate. Increased resistance to antibiotics in the

presence of salicylate has been attributed to the alteration in membrane-associated

proteins such as porins and transporters, leading to a reduction in drug accumulation [7].

This salicylate-associated phenotypic resistance is non-heritable, and results in reduced

susceptibility to mechanistically and structurally distinct antimicrobials [26-37].

Phenotypic resistance has been attributed to a weak acid effect, which is thought to be

due to an increase in membrane potential and altered permeability and a decrease in

internal pH, the pH gradient, and the proton motive force [38]. Foulds et al. [39]

demonstrated a 3- to 5-fold decrease in permeation of cephalosporins through the outer

membrane of E. coli induced with salicylate.

Salicylate has been shown to alter energy metabolism in many organisms. For

example, in E. coli, growth with salicylate altered the expression of over 130 genes [41].

In addition, it was shown to dissipate the proton gradient across the inner membrane,

chelate iron, induce heat shock as well as inhibit growth [40-41]. Similarly, in Bacillus

subtilis, salicylate was shown to impair energy metabolism as observed through a down-

regulation of ATPases, suggesting energy impairment [41]. Salicylate also induced the

general stress sigma factor B (sigB), and sigB dependent genes in B. subtilis. In addition,

salicylate decreases metabolism of purines, pyrimidines, coenzymes, as well as

metabolism of carbohydrates involved in glycolysis [42]. Growth with salicylate has

similar modulatory effects on metabolism in S. aureus. Therefore, it is of no surprise that

   4  

salicylate altered metabolism in S. aureus, resulting in inhibition of glycolysis as seen in

the downregulation of glyceraldehyde-3-phosphate dehydrogenase (gapA2) and

phosphoglucoisomerase (pgi) [43]. The findings by Riordan et al. [43] indicate an

impairment of growth and overall alteration in central and energy metabolism. This

existing data suggest that salicylate has pleiotropic effects on bacterial cells that are

physiochemical and metabolic.

1.4 Staphylococcus aureus

Staphylococcus aureus is a low GC Gram-positive, non-spore forming bacteria

that was discovered in the 1880s [44]. S. aureus is a facultative anaerobe [46] that can

grow at temperatures between 25°C to 43°C and at pH levels of 4.8 to 9.4 [45]. S. aureus

mainly colonizes the membranes and skin of warm-blooded animals, and infections range

from benign skin lesions to life-threatening systemic illnesses such as endocarditis and

osteomyelitis [45]. The Center for Disease Control and Prevention estimated in 2005 that

there were 31.8 culture confirmed invasive methicillin resistant S. aureus (MRSA)

infections in the U.S. per 100,000 individuals, which amounted to 94,360 cases [47]. The

primary mode of transmission is through direct contact, usually skin-to-skin contact with

a colonized or infected individual, although contact with contaminated objects or surfaces

also plays a role [48-51]. S. aureus is capable of producing many toxins and is able to

acquire resistance to many antibiotics [45]. Currently, greater than 60% of S. aureus

isolates are resistant to methicillin and some strains have developed resistance to more

than twenty antimicrobial agents [52]. Acquisition of resistance in S. aureus commonly

results from either gene mutations leading to drug target modifications or reduction of

   5  

drug efficacy, as well as acquisition of a resistance gene(s) from other organisms by

some form of horizontal genetic exchange [53]. The genome of S. aureus is

approximately 2.8 Mbp with a GC content of 33% [45, 52, 53].

1.5 Antibiotic Resistance in S. aureus

Staphylococcus aureus is exceptionally resistant to a wide-range of antibiotics

(for review see [44]). This resistance evolved in S. aureus via horizontal gene transfer,

chromosomal mutation and antibiotic selection [44]. Antibiotic resistance in S. aureus

emerged in a series of waves [44]. The first wave of resistance began in the mid 1940s

where penicillin-resistant strains began to surface in hospitals shortly after its

introduction [54, 55]. Introduction of methicillin in 1960 initiated the second wave of

resistance. Methicillin is a narrow spectrum β-lactam antibiotic that inhibits cross-

linkages between the linear peptidoglycan polymer chains that make up a major

component of the Gram positive cell wall [56]. It was as early as 1961 that resistance to

methicillin was detected in S. aureus. The mecA gene, encoding alternative penicillin

binding protein 2 (PBP2) responsible for the methicillin resistance phenotype, was not

identified until more than twenty years later. The emergence of methicillin resistant S.

aureus (MRSA) strains led to an increase in the use of vancomycin [57, 58]. Vancomycin

is a glycopeptide antibiotic used in the prophylaxis and treatment of serious infections

caused by Gram positive bacteria [57, 58]. Vancomycin acts by inhibiting proper cell

wall synthesis through formation of hydrogen bond interactions with the terminal D-ala-

D-ala moieties of the NAG/NAM peptides [57, 58]. The first vancomycin intermediate

resistant S. aureus (VISA) strain was reported in 1997 with MICs <16 µg/ml [59]. The

   6  

first VRSA strain was isolated in June 2002 with high-level vancomycin-resistance and

an MIC=1024 µg/ml [59]. This isolate was also found to be resistant to aminoglycosides,

rifampin, and tetracycline. This VRSA strain was later determined to acquire the

vancomycin resistance gene through horizontal gene transfer from Enterococcus faecalis

[57, 58]. The invasion of MRSA into the community constitutes the fourth and latest

wave of resistance. It is during this wave that community associated (CA) MRSA and

vancomycin resistant strains began to emerge [44]. Despite all the resistance only two

antibiotics besides vancomycin, linezolid and daptomycin have been approved for

therapy since the 1990s [60].

1.6 Salicylate associated phenotypic and genotypic antibiotic resistance in S. aureus

Phenotypic antibiotic resistance was first characterized in E. coli. Resistance to

chloramphenicol and ampicillin was induced during incubation with the weak acids

acetate or benzoate, the NSAIDs aspirin and salicylate, and other chemical repellants

such as dimethyl sulfoxide (DMSO), and 1-methyl-2-pyrrolidinone, [29]. Importantly,

cells were sensitive to the antibiotics when grown in the absence of these inducers, and

thus resistance was described to be inducible and non-heritable. Salicylate-inducible

antibiotic resistance in E. coli was subsequently found to be due, in part, to increased

transcription of the marRAB operon [25]. The marRAB operon consists of marR, which

encodes a negative regulator of the operon, and marA, a transcriptional activator.

Salicylate has been found to interact with the ligand binding domain of MarR, interfering

with its ability to efficiently recognize the mar operator, marO. This, in turn, leads to

induction of MarA and a decrease in antibiotic accumulation by reduced production of

   7  

outer membrane porins OmpF and OmpC, and a concomitant increase in the production

of the multidrug efflux pump AcrAB [61, 62]. Salicylate has also been shown to induce

phenotypic antibiotic resistance in Salmonella typhimurium. When grown in the presence

of salicylate, S. typhimurium developed increased resistance to chloramphenicol, and

enoxacin. This was found to be due to the induction of a S. typhimurium mar system

homologous to the mar operon of E. coli [28]. Klebsiella pneumoniae also exhibits

increased phenotypic resistance to tetracycline, β-lactams, clindamycin and norfloxacin

in the presence of salicylate, which has been reported to be due to increased transcription

of ramA, encoding a MarA homolog [35].

Phenotypic antimicrobial resistance is also induced by salicylate in S. aureus.

When exposed to salicylate, S. aureus develops increased resistance to fluoroquinolones,

the steroid antibiotic fusidic acid, hard surface disinfectants and ethidium bromide [26,

33]. The mechanism for phenotypic resistance in S. aureus is thought to be achieved in

part by the upregulation of drug efflux pumps including NorA, NorB, MdeA and SepA

[7, 63-66] and through alterations in membrane permeability [33, 43, 67-71]. As in other

bacteria, phenotypic resistance in S. aureus appears to be dependent on Mar family

homologs, such as the Sar-family of proteins, MgrA, as well as the general stress

resistance protein sigma factor B [66, 72-76]. Sigma factor B is one of two alternative

sigma factors in S. aureus [77] and is essential for the general chemical and physical

stress response of the organism [78]. sigB regulation has been found to be intertwined

with the expression of SarA, which regulates the expression of a number of

staphylococcal virulence factors [78-81]. Analyses by Riordan et al. [71] determined that

altered expression of sigB and sarA is not required for the salicylate-inducible

   8  

mechanism. Riordan et al. [43] also observed impairment in the glycolytic pathway as

seen through a decrease in glyceraldehyde-3-phosphate dehydrogenase (gapA2) and

phosphoroglucoisomerase (pgi) in response to salicylate stress, which is supported by

evidence in eukaryotes [8, 9, 82-84].

Salicylate has been shown to alter the level of resistance as well as the frequency

at which antibiotic resistance occurs in S. aureus [7, 26, 33, 85]. Specifically, salicylate

has been observed to increase the frequency at which S. aureus mutates to become

resistant to fluoroquinolones and fusidic acid [7, 26, 33, 85]. The addition of salicylate

significantly increased the number of ciprofloxacin and norfloxacin resistant S. aureus

colonies compared to resistant colonies selected in the absence of salicylate;

ciprofloxacin-resistant mutants arose at mutation frequencies of 1.8 x 10-9 on plates

containing ciprofloxacin, compared to a mutation frequency of 1.8 x 10-7 on plates

containing ciprofloxacin and salicylate [33]. Colonies selected from ciprofloxacin in the

presence of salicylate containing plates had MICs > 0.8 mg/l, which was higher than

colonies selected in the absence of salicylate [33]. The mutations leading to

fluoroquinolone and fusidic acid resistance in these isolated were heritable, and

resistance to these antibiotics occurred at unrelated loci within the S. aureus genome [86,

87]. The underlying mechanism of salicylate-associated genotypic resistance [26, 33] and

the extent to which salicylate alters resistance to other antibiotics is currently unknown.

1.7 Hypothesis and aims of the study

This study seeks to understand the basis of salicylate-associated genotypic

antibiotic resistance (SAGAR) in the deadly human pathogen, S. aureus. The overarching

hypothesis is that the SAGAR phenotype is common to mechanistically and structurally

   9  

distinct antibiotics, and is attributable to the effect(s) of salicylate on S. aureus

metabolism. The following aims test this hypothesis:

Aim 1. Assess the ability of salicylate to alter the frequency at which genotypic

resistance to mechanistically and structurally distinct antibiotics occurs in S. aureus.

Aim 2. Examine the role for salicylate-associated metabolic stress in the genotypic

antibiotic resistance phenotype.

   10  

CHAPTER TWO: CHARACTERIZATION OF THE SALICYLATE-

ASSOCIATED GENOTYPIC ANTIBIOTIC RESISTANCE PHENOTYPE IN S.

AUREUS

2.1 Background

The nonsteroidal anti-inflammatory drug (NSAID) salicylate has been shown to

increase the frequency at which heritable (genotypic) fusidic acid and ciprofloxacin

resistance occurs in S. aureus [7, 26, 33, 43, 69, 70]. These antibiotics have distinct

cellular targets, and are structurally unique. This salicylate-associated antibiotic

resistance (SAGAR) suggests that salicylate, and perhaps other NSAIDs, may have a

generalized effect on mutation frequency in the cell. Yet, salicylate has not been shown

to be directly mutagenic by the Ames test [85], which suggests that this increase in

mutation frequency may result from the alteration of an existing physiological process

which occurs in combination with specific antibiotic chemistries or antibiotic mechanistic

activities. For example, bactericidal drugs such as ciprofloxacin have been shown to

increase oxidation of NADH via the electron transport chain [88, 89], resulting in

formation of the reactive oxygen species (ROS) superoxide [90-92]. Superoxide and

other ROS damage iron-sulfur clusters, making ferrous iron available for oxidation by the

Fenton reaction [88, 89]. The Fenton reaction leads to hydroxyl radical formation, which

damages DNA, proteins and lipids, ultimately resulting in cell death [88, 89].

Subinhibitory concentrations of certain antibiotics, specifically compounds whose

   11  

primary mode of action is DNA damage, are known to enhance mutation rates in bacteria

[93, 94]. This elevation in mutation frequency is partly a result of transcriptional changes

in genes responsible for DNA repair and preservation of the integrity of the genome [93,

94]. This effect is also observed with rifampin, an antibiotic that interacts specifically

with the β subunit of the bacterial RNA polymerase encoded by the rpoB gene [95].

Growth with salicylate also leads to a non-heritable (phenotypic) increase in

resistance to many antimicrobials [7, 26, 33, 43, 69-71, 96]. As in E. coli, this salicylate-

inducible phenotypic resistance in S. aureus partly results from a decrease in drug

accumulation due to alterations in membrane permeability, proton motive force, and

efflux [7, 25, 26, 33, 36, 37, 43, 69-71, 96]. A number of proteins have been determined

to be involved in this phenotypic resistance mechanism of S. aureus including: multidrug

efflux pumps NorA, NorB, MdeA, and SepA as well as other chromosomally encoded

efflux pumps [7, 63-66, 97]; the global regulatory protein MgrA [66, 73, 76];

staphylococcal accessory regulator (SarA) [74]; and alternative sigma factor B (SigB)

[70]. Mutations in the S. aureus genes encoding these proteins resulted in increased [70,

74, 98-100] or decreased [63-65, 80, 97, 99] susceptibility to antimicrobials. Phenotypic

resistance may allow the cell prolonged exposure to low levels of antibiotic, leading to

the acquisition of mutations and high level heritable (genotypic) resistance [101].

Currently, the salicylate-associated genotypic antibiotic resistance (SAGAR)

phenotype has only been described for a limited number of antibiotics, or for antibiotics,

which are not commonly indicated for S. aureus infections. In addition, the relationship

between salicylate-inducible phenotypic resistance and the SAGAR phenotype is

unknown. The following experiments are designed to examine the SAGAR phenotype,

   12  

and to test the hypothesis that growth of S. aureus with salicylate increases the frequency

at which resistance to structurally and mechanistically distinct antibiotics occurs.

2.2 Methods

Bacterial strains and culture media and conditions

S. aureus strains used in this study are listed in Table 2.1. Unless otherwise noted,

strains were grown aerobically at 37°C with shaking (200 RPM) in baffled Erlenmeyer

flasks (5:1 volume ratio of flask:media). Cultures were generally maintained in tryptic

soy broth (TSB) or TSB with 1.5% agar (TSA), and stocked at -80°C in TSB with the

addition of 20% (vol/vol) glycerol.

Antibiotics and NSAIDs

All antibiotics and nonsteroidal anti-inflammatory drugs (NSAIDs) were

dissolved, filter sterilized and stored according to Material Safety Data Sheet guidelines

(Version 3 [102]). The antibiotics used in this study included ciprofloxacin (2.5 mg/ml in

0.1 N HCl), norfloxacin (5 mg/ml in 0.1 N HCl), fusidic acid (1 mg/ml in sterile water),

vancomycin (1 mg/ml in sterile water), tetracycline (1 mg/ml in sterile water), rifampin

500 µg/ml in methanol), oxacillin (1 mg/ml in sterile water) and lincomycin (1 mg/ml in

sterile water). The NSAIDs and weak acids used in this study included: sodium salicylate

(0.5 M in sterile water), acetylsalicylic acid (3 mM in sterile water) and sodium benzoate

(0.5 M in sterile water).

   13  

Table 2.1. Bacterial Strains and Plasmids

Strain name Relevant Characteristics Source/reference S. aureus SH1000 rsbU+ derivative of 8325-4 Dr. Lindsey Shaw, USF, [103] SaRNH-1 SH1000Δ sigB Dr. Lindsey Shaw, USF, [103] Newman Wild-type Dr. John Gustafson, NMSU, [104] SaRNH-2 Newman ΔmgrA Dr. John Gustafson, NMSU, [105] LAC USA300 CA-MRSA Dr. Lindsey Shaw, USF, [106] E. coli MG1655 Wild-type Dr. James Riordan, USF, [107]

Table 2.2 Primers used in this study Primer name Sequence (5’à3’) grlA+2402 ACTTGAAGATGTTTTAGGTGAT grlA+2961 TTAGGAAATCTTGATGGCAA grlB+1520 CGATTAAAGCACAACAAGCAAG grlB+1894 CATCAGTCATAATAATT CTC gyrA+2311 AATGAACAAGGTATGACACC gyrA+2533 TACGCGCTTCAGTATAACGC gyrB+1400 CAGCGTTAGATGTAGCAAGC gyrB+1650 CCGATTCCTGTACCAAATGC

All primers were designed based on this study. Determination of mutation frequency

The protocol for mutation frequency to antibiotic resistance was adapted from

Foster, 2006 [108]. Three independent cultures were grown under standard conditions in

TSB overnight before sampling 0.1 ml onto TSA alone, TSA with antibiotic at ½ X, 1X,

and 2X MIC, TSA with NSAID, or TSA with antibiotic and NSAID. Plates were

incubated for 24 h before counting colony forming units (CFU/ml). Mutation frequency

(µ) and fold-change in mutation frequency was calculated from colony counts using the

following equations:

   14  

Eq. 2.1:  µμ  antibiotic =CFU  (Antibiotic)CFU  (TSA)

Eq. 2.2:  µμ  NSAID =CFU   Antibiotic+ NSAID

CFU  (NSAID)

Fold change was calculated using the following equation:

Eq. 2.3:  Fold  Change =  µμAntibiotic+ NSAID

Antibiotic Differences in the mutation frequency between control and treatment cultures were

compared using a t-test (α=0.05, n≥3) (R).

Antibiotic susceptibility by the minimum inhibitory concentration assay (MIC)

MICs were determined as described [109] with slight adaptations. Overnight S.

aureus MHB cultures were diluted to an OD600 = 0.01 in fresh MHB. A stock of the

desired antibiotic was prepared at two times the highest concentration in MHB. Serial 2-

fold dilutions of respective drugs were prepared in MHB. Of the diluted culture, 1 ml was

added to each tube to achieve the final concentration with a final volume of 2 ml. A

negative control containing the growth medium plus antibiotic was prepared. A positive

control containing the growth medium and culture was also prepared. The set of tubes

were incubated static overnight at 37°C. The MIC for each independent sample was

recorded, as the lowest concentration of antibiotic at which there was no visible growth.

S. aureus chromosomal DNA extraction

A single colony of S. aureus SH1000 was grown under standard conditions

overnight. The following day the cells were pelleted by centrifugation at 3,700 x g for 10

   15  

min. Pelleted cells were resuspended in 5 ml 1X TE buffer (pH 8.0) and centrifuged as

above. Cell pellets were then resuspended in 600 µl of 1X TE buffer and transferred to a

1ml Eppendorf tube containing 0.5 cm3 of 0.1 mm glass beads. The tube was placed in a

bead beater (Mini-bead beater-16, Biospec) and pulsed at 10 sec intervals for a total of 60

seconds without a break. Homogenates were then centrifuged for 5 min at 13,200 x g.

The supernatant was transferred to a sterile microtube, and 0.2 ml of 1.6% (vol/vol)

sarkosyl and a total of 25 µg of proteinase K was added to the tube and incubated at 60°C

for 60 min. Eight-hundred microliters of phenol/chloroform/isoamyl was added,

vortexed, and centrifuged at 13,200 x g for 5 min. The upper aqueous layer was

transferred to a fresh Eppendorf tube, and 0.5 ml isopropyl alcohol and 100 µl of 3 M

sodium acetate were added and mixed by inversion. This was allowed to incubate at -

80°C for 15-30 min. The samples were then centrifuged at 13,200 x g for 5 min, and the

supernatant was carefully discarded. Five-hundred microliters of 70% ethanol was added

to the pellet and centrifuged at 13,200 x g for 5 min. The supernatant was then discarded.

The samples were allowed to air dry for 3-4 min at room temperature with the lid open.

Two-hundred microliters of ddH2O was used to resuspend the DNA.

Sequencing of target site modifications in antibiotic resistant mutants

Antibiotic resistant and susceptible isolates were passaged three times in the

absence of antibiotic before MICs were determined. DNA from two representative

isolates was extracted and PCR was used to amplify a 559-bp fragment of grlA using

primers grlA+2402/grlA+2961, a 374-bp fragment of grlB using primers

grlB+1520/grlB+1894, a 222-bp fragment of gyrA using primers gyrA+2311/gyrA+2533,

   16  

and a 250-bp fragment of gyrB using primers gytB+1400/gyrB+1650 (Table 2.2). grlAB

and gyrAB were amplified at an annealing temperature of 55°C, and an extension time of

50 seconds for 30 cycles. Products were confirmed by ethidium bromide agarose gel

electrophoresis and were purified using Qiagen purification kit (Qiagen, Valencia, CA)

per the manufacturer’s instructions. Products were sequenced using standard Sanger dye

chain-termination sequencing through the services of MWG Operon (Huntsville, AL)

using forward and reverse primers (Table 2.2) to read both template and coding strands.

Sequencing reads were compared to publically available sequences of S. aureus N315

(NCBI) by pairwise alignment. Target site modifications were recorded and annotated

according to genomic position.

Selection for resistance to sodium salicylate

Spontaneous selection

An overnight culture of S. aureus strain SH1000, LAC and E. coli strain MG1655

(K-12) was prepared under standard conditions. Large TSA plates (100 x 200 mm) as

well as TSA gradient plates were prepared at ¼, ½, 1, 2, 3, 4, 5, 6 X MIC of either

salicylate or acetyl salicylic acid. Cultures (100 µl) were inoculated onto each of the

plates incubated at 37°C, and putative resistant colonies were selected after 24 h.

Colonies were then passaged twice in the absence of NSAID before MICs were

performed to determine if the recovered isolates had a higher level of resistance to the

NSAID when compared to WT.

   17  

N-methyl-N-nitro-N-nitrosoguanidine (NTG)-mutagenesis

To select for a salicylate resistant straing NTG mutagenesis was used. A single

colony of S. aureus SH1000 was prepared under standard conditions. One milliliter of

overnight culture was inoculated into 100 ml TSB and allowed to grow for 3 hours before

addition of NTG (50 µg/ml final) or an equal volume of sterile water (control). Cell

counts (CFU/ml) were performed immediately before addition of NTG. Cells were grown

for 45 min with and without NTG before harvesting by centrifugation at 1,929 x g for 10

min. NTG has been shown to induce at least one mutation per cell under the above

growth conditions, which has been shown to correlate with a 50% survival rate [110].

NTG is known to add alkyl groups to O6 of guanine and O4 of thymine [110]. The

supernatant of NTG-treated and non-treated (control) cultures was then removed and

discarded, and the cells were resuspended in 100 ml of TSB by vortexing. These cells

were then centrifuged again as before and the supernatant was discarded before being

resuspended in 100 ml of TSB and allowed to grow for 2 hours. Cultures were then

sampled to verify the efficiency of NTG killing, and serial dilutions were plated to

determine final CFU/ml for NTG-treated compared to untreated cultures. Multiple

libraries of NTG mutants were stocked by taking ten 1 ml aliquots in Eppendorf tubes

and centrifuging them at 1,372 x g for 3 minutes. The supernatant was discarded and the

cells were resuspended in TSB with 20% (v/v) glycerol and stored at -80°C.

Stepwise selection

An overnight culture of S. aureus strain SH1000 was prepared under standard

conditions. Cells were inoculated in fresh TSB at 1:100 containing 1.5 mg/ml salicylate.

   18  

The following day cells were plated on to TSA plates for cell counts and were also

inoculated into fresh TSB at 1:100 containing 2X the initial concentration of salicylate.

This process was repeated until no growth was recovered on TSA plates. MICs were

determined for all colonies recovered following passage in TSB without antibiotic to

determine if any of the cells acquired a higher level of resistance.

2.3 Results

Antibiotic specificity of the salicylate-associated genotypic resistance phenotype

Salicylate-associated genotypic antibiotic resistance (SAGAR) has been described

for the fluoroquinolone antibiotics ciprofloxacin and norfloxacin, as well as the steroid

antibiotic fusidic acid in both laboratory and MDR strains of S. aureus. To further assess

the scope of this phenotype, the impact of salicylate on the frequency of resistance to a

spectrum of antibiotics belonging to several drug classes was investigated for S. aureus

laboratory strain SH1000, and when possible, the CA-MRSA strain LAC (Table 2.1). Of

the nine antibiotics tested, salicylate was observed to only alter the frequency at which

resistance to ciprofloxacin (CipR), norfloxacin (NorR) and lincomycin (LinR) occurred

(Table 2.3). The frequency at which CipR occurred in SH1000 increased by 27-fold

(p=0.03) in the presence of salicylate (Table 2.3). The frequency at which CipR mutants

in strain LAC occurred with salicylate also increased, but only by 6.2-fold and not

significantly. Growth of SH1000 with salicylate also significantly increased the mutation

frequency to NorR by 4-fold (p=0.01). Unlike ciprofloxacin and norfloxacin, the

frequency at which resistance to lincomycin occurred with salicylate decreased by 3-fold

in SH1000, but not significantly (p=0.05). Mutation frequency to norfloxacin and

   19  

lincomycin resistance for strain LAC was not determined, due to its high intrinsic

resistance to these antibiotics; MICs for ciprofloxacin and lincomycin in LAC were 64

µg/ml and an MIC of 25 µg/ml, respectively.

To determine the effect of adapting S. aureus to salicylate on the SAGAR

phenotype, cultures were grown overnight with salicylate (adaptive environment) or

without (un-adapted) and then tested for SAGAR on plates, which contained

ciprofloxacin alone, or ciprofloxacin and salicylate. Adaptation to salicylate did not

influence the frequency at which CipR colonies occurred compared to un-adapted cultures

when selected on ciprofloxacin plates without salicylate (Fig. 2.1). Also, adaptation to

salicylate had no effect on the increase in frequency to CipR observed when selected in

the presence of salicylate. Collectively, these findings reveal that the SAGAR phenotype

is antibiotic specific. Furthermore, the results suggest that salicylate is not a chemical

mutagen in S. aureus, agreeing with previous studies in Salmonella typhimurium [28],

and reveals that adaptation of cultures to salicylate has no apparent impact on SAGAR.

   20  

Table 2.3. Antibiotic specificity of salicylate-associated genotypic phenotype.

Antibiotic Strain Plated (µg/ml)

Mutation Frequency Fold

change SD Antibiotic (Abx)

Abx + 500 µg Salicylate

Ciprofloxacin

SH1000 1 4.9 X 10-8 8.5 X 10-7 27* 2.9 X 10-7

SaRNH-1 1 5.2 X 10-7 2.1 X10-6 8.5* 6.9 X 10-7

LAC Newman SaRNH-2

16 1 1

8.2 X 10-4

3.2 X 10-6

3.8 X 10-6

2.4 X 10-3

4.7 X 10-6

4.3 X 10-6

6.2 1.5 1.1

3.6 X 10-3

1.2 X 10-6 1.8 X 10-6

Fusidic Acid SH1000 0.5 6.1 X 10-7 8.7 X 10-7 1.4 6.0 X 10-7

SaRNH-1 2 3.6 X 10-6 3.2 X 10-6 0.9 2.4 X 10-6

LAC 4 7.7 X 10-8 1.6 X 10-7 2.2 1.2 X 10-7

Lincomycin SH1000 1 3.4 X10 -7 1.3 X 10-7 0.4* 1.0 X 10 -7

SaRNH-1 1 1.2 X 10-7 1.2 X 10-9 0.02* 5.9 X 10 -10

Norfloxacin

SH1000 4 1.5 X 10-7 6.6 X 10-7 5.2* 2.6 X 10-7

SaRNH-1 Newman SaRNH-2

4 2 2

6.8 X 10-8

2.1 X 10-6

2.7 X 10-6

6.9 X 10-7

4.6 X 10-6 4.3 X 10-6

7.2 2.8 1.6

6.7 X 10-7

1.6 X 10-6

1.3 X 10-6

Rifampin SH1000 1 2.9 X 10-7 2.3 X 10 -7 0.8 1.4 X 10-7

SaRNH-1 1 6.2 X 10-8 6.3 X 10-8 1 3.1 X 10-8

LAC 0.1 9.8 X 10-8 1.2 X 10-7 1.7 8.6 X 10-8

Tetracycline SH1000 0.5 5.6 X 10-7 1.0 X 10-6 2 3.8 X 10-7

SaRNH-1 190 4.5 X 10-6 7.0 X 10-6 1.9 1.7 X 10-6

Vancomycin SH100 3 5.7 X 10-6 2.3 X 10-6 0.3 3.0 X 10-6

LAC 7 6.43 X 10-8 2.6 X 10 -8 0.4 1.9 X 10-8

SD indicates standard deviation for mutation frequency with antibiotic (abx) with salicylate. Asterisks denotes statistical significance p< 0.05, n=3. Strains were plated at 1 or 2X the MIC.

   21  

Figure 2.1. Frequency of resistance to ciprofloxacin in S. aureus cultures adapted to salicylate. CFU/ml of CipR colonies of SH1000 are plotted as a function of treatment: growth overnight with salicylate (adaptive environment) or without (control, non adaptive) and then plated to ciprofloxacin alone (filled bars) or ciprofloxacin and salicylate (hatched bars). Asterisks denote statistical significance by ANOVA and Tukey’s HSD (p= 0.01).

Antibiotic specificity of the SAGAR phenotype

Broth microdilution minimum inhibitory concentration (MIC) assays were used to

determine if selection for resistance to antibiotics with salicylate conferred the same level

of resistance as selection in the absence of salicylate. MICs for antibiotic resistant

isolates of strains SH1000, SaRNH-1 (a.k.a. SH1000sigB::tet), and LAC exceeded

clinical and laboratory standards institute (CLSI) breakpoints for all drugs, except for

lincomycin (breakpoint = 10 µg/ml) and vancomycin (breakpoint = 16 µg/ml) (Table

   22  

2.4). MICs for ciprofloxacin did not differ for resistant isolates selected with or without

salicylate, but were lower for fusidic acid, lincomycin and norfloxacin resistant isolates

for all strains when selected in the presence of salicylate (Table 2.4). MICs for SaRNH-1

resistant isolates did not differ between MICs for SH1000 resistant isolate strains for

ciprofloxacin. Interestingly, a difference in the level of resistance between resistant

isolates was observed. For example SaRNH-1 lincomycin isolates selected in the absence

of salicylate had higher MICs (4 µg/ml) than SaRNH-1 lincomycin isolates selected in

the presence of salicylate with an MIC of 2 µg/ml, These results suggest that the

mutations, which lead to resistance to some antibiotics when selected in the presence of

salicylate, may differ from those that lead to resistance in the absence of salicylate. Also,

a difference in colony morphology was observed. Small colonies like variants were

observed on antibiotic plates selected in the presence and absence of salicylate.

   23  

Table 2.4. Minimum inhibitory concentrations (MICs) for antibiotic resistant isolates selected with or without salicylate

Drug Strain Initial MIC (µg/ml)

MICs (µg/ml) Sal - Sal + LC SCV LC* SCV*

Ciprofloxacin SH1000 0.5 4 4 4 4

SaRNH-1 1 4 4 4 1 LAC 64 512 512 124.3 124.3

Fusidic Acid SH1000 0.24 124.8 124.8 124.8 124.8

SaRNH-1 0.98 NP NP NP NP LAC 0.24 249.8 124.9 NP NP

Lincomycin SH1000 1 4 4 4 4 SaRNH-1 1 4 4 2 2

Norfloxacin SH1000 2 16 16 8 4 SaRNH-1 4 32 16 16 8

Rifampin SH1000 0.25 NP NP NP NP

SaRNH-1 0.5 NP NP NP NP LAC 0.02 NP NP NP NP

Tetracycline SH1000 1 4 1 4 1 SaRNH-1 64 NP NP NP NP

Vancomycin SH100 1 4 4 4 4 LAC 3 NP NP NP NP

All values are in micrograms per milliliter. Large colony variant (LC); small colony variant (SC). NP denotes not performed. Sal denotes salicylate at 500 µg/ml

Characterization of mutations conferring resistance to fluoroquinolones

Quinolone resistance is gained through modification of gyrAB and grlAB targets

as well as modification through the norA efflux pump promoter sequence, which are

associated with clinical levels of resistance [111-114]. A difference in the level of

quinolone resistance in some strains selected for in the presence or absence of salicylate

was detected. It was of interest to sequence gyrAB and grlAB quinolone resistance

determining region (QRDR) to determine if there was a difference between the acquired

   24  

SNPs. All mutations conferring resistance to ciprofloxacin and norfloxacin were

determined to be in a 347-bp region of grlA, encoding the A subunit of topoisomerase IV.

This resulted in a common mutation at amino acid Ser80 [113, 115], as well as less

common (or unreported) mutations at Arg43 and Ala115 (Table 2.5). The location of the

SNP had no apparent effect on the level of ciprofloxacin resistance, MIC = 4 µg/ml

(Table 2.4). However, for norfloxacin resistant large colony (LC) isolates, the location of

the SNP did not differ, but the level of resistance did: MIC of 16 µg/ml (without

salicylate) and 8 µg/ml (with salicylate) (Table 2.4 and 2.5). Lincomycin, a macrolide,

acts on the 50S ribosomal subunit, specifically targeting 23S rRNA [116]. Macrolides,

such as erythromycin have been shown to target A2058G/U or A2059G, however, the

exact mutation responsible for conferring lincomycin resistance remains to be identified

[117]. Therefore, target site modifications conferring resistance to lincomycin were not

determined.

Table 2.5. Sequencing results for grlA in fluoroquinolone resistant isolates in SH1000 compared to NCBI S. aureus sp N315.

SH1000 Antibiotic Original SNP Genomic Mutation Original AA Location Location

LC Cipro

A G 1,356,325 His Arg 43 LC* C T 1,356,564 Phe Ser 80 LC C G 1,356,671 Ala Ala 115 LC, LC* Nor C T 1,356,564 Phe Ser 80 SCV C T 1,356,672 Val Ala 115 Large colony variant (LC); small colony variant (SC). Selection for antibiotic resistance in the presence of salicylate denoted by *. Single nucleotide polymorphism (SNP). Amino acid location (AA location).

   25  

Role of sigB and mgrA in the SAGAR phenotype

Alternative sigma factor B (σB) directs the transcription of more than one hundred

genes in response to different stressors [98, 118]. Sigma factor B is necessary for full

expression of salicylate-inducible phenotypic resistance, and its expression was shown to

be upregulated upon salicylate exposure [71, 74]. MgrA is a helix-turn-helix DNA

binding protein and, like SigB, regulates many S. aureus genes (approx. 355) [66]

including genes shown to control autolytic activity and the expression of several

virulence factors, including alpha-toxin, nuclease and protein A [66]. Importantly, MgrA

negatively regulates the multiple drug efflux pumps NorA, NorB, NorC, Tet38 and

AbcA, shown to be important for resistance to fluoroquinolones [75, 76], and mgrA

transcript levels are repressed by salicylate [71]. In previous studies, an mgrA mutant

showed resistance to ciprofloxacin, which was linked to increased expression of norA

[66, 70]. Since σB and MgrA contribute to the phenotypic resistance mechanism, we

aimed to ascertain the role for these regulators in the SAGAR phenotype using strains,

which are isogenic for either sigB or mgrA (Table 2.1).

Removal of sigB in SH1000 (SH1000sigB::tet) did not significantly alter the

mutation frequency in the absence of salicylate to ciprofloxacin (p=0.07), norfloxacin

(p=0.43) or lincomycin (p=0.09) when compared to WT SH1000 (Table 2.3). However,

removal of sigB significantly reduced the mutation frequency to ciprofloxacin with the

addition of salicylate from 27-fold in SH1000 down to 8.5-fold in SH1000sigB::tet

(p=0.02). Interestingly, the same effect was not seen with norfloxacin, where the

mutation frequency in SH1000sigB::tet slightly but insignificantly increased to 7.2-fold

compared to 5.2-fold in SH1000 (p=0.09). Although SAGAR was not observed for

   26  

lincomycin in SH1000, in SH1000sigB::tet salicylate decreased the mutation frequency

to lincomycin by 0.02-fold (or a 50-fold reduction) (p=0.04). These findings emphasize a

significant role for sigB in the salicylate associated genotypic antimicrobial resistance

phenotype (SAGAR) phenotype.

For mgrA analysis, S. aureus strain Newman [110], and its mgrA null derivative

(Table 2.1) were used to assess the potential contribution of mgrA to the SAGAR

phenotype. In Newman, salicylate was shown to slightly increase mutation frequency to

ciprofloxacin and norfloxacin (p=0.02 and p=0.01, respectively). However, the increase

in ciprofloxacin mutation frequency was less than that observed for strains SH1000 or

LAC (Table 2.3). Deletion of mgrA in the Newman background did not significantly alter

mutation frequency in the presence of salicylate for ciprofloxacin (p=0.10) and

norfloxacin (p=0.07) (Fig. 2.2). Although we did not observe a significant change in

mutation frequency upon deletion of mgrA in the Newman background, we cannot rule

out a potential role for mgrA in SH1000 as our results have indicated strain specificity for

SAGAR.

Investigating the chemical signature associated with salicylate associated genotypic

resistance to antibiotics

Salicylate-inducible phenotypic resistance to fluoroquinolones has been attributed

to the carboxylic acid group of salicylate and acetyl salicylic acid [33], however the

importance of this functional group, and others intrinsic to salicylate in SAGAR, is

unknown. To identify chemical features of salicylate that contribute to SAGAR, the

effect of the salicylate analogs sodium benzoate (weak acid) and acetylsalicylic acid

   27  

(NSAID) (Fig. 2.2) on mutation frequency to ciprofloxacin, norfloxacin and lincomycin

resistance was determined.

Mutation frequency to ciprofloxacin resistance was significantly higher with

salicylate when compared to sodium benzoate, acetylsalicylic acid (ASA), and controls

(p=0.03) (Fig. 2.3). Benzoate and ASA differ structurally at the ortho position, the

hydroxyl group of salicylate being reduced to hydrogen at this position in benzoate,

whereas in ASA, this hydroxyl group is acetylated (Fig. 2.2). For lincomycin, a

significant decrease in mutation frequency was observed for benzoate, salicylate, and

ASA when compared to untreated cultures (p<0.05), but not between treatments (Fig.

2.3). Similarly, for norfloxacin, there was no significant difference in the mutation

frequency of SH1000 to ciprofloxacin with salicylate when compared to benzoate, or

salicylate compared to ASA, or benzoate to ASA (p=0.07, p=0.06, and p=0.05,

respectively). Our results indicate that salicylate is needed to fully propagate the SAGAR

phenotype for ciprofloxacin and norfloxacin.

Figure 2.2. Structural differences between salicylate, benzoate and acetyl salicylic acid. Box indicates altered chemistry at the ortho-position.

   28  

Figure 2.3. Dependence on salicylate chemical structure for SAGAR. Fold change in frequency of resistance to antibiotics determined for cultures selected in the presence of salicylate (black), benzoate (hatched), and acetyl salicylic acid (dotted). Error bars indicate standard deviation of the mean, and asterisks denote groups that differ significantly by ANOVA and Tukey’s HSD (a=0.05, n=3).

2.4 Discussion

The results of this study revealed that the SAGAR phenotype for salicylate is drug

specific. The salicylate associated antibiotic resistance (SAGAR) phenotype was only

observed for the fluoroquinolones, ciprofloxacin and norfloxacin, in strain SH1000. The

phenotype was more prominent with ciprofloxacin than with norfloxacin. This suggests

that SAGAR is highly selective and sensitive to subtle changes in antibiotic structure.

Ciprofloxacin and norfloxacin are structurally very similar in that both have fluorinated

quinoloic acid cores; however there are differences. For example, ciprofloxacin has a

cyclopropane ring at the N-1 position, while norfloxacin has an ethyl group at the same

position [119]. In addition, studies by Chin et al. [120] found that ciprofloxacin was 4 to

   29  

32 times more active than norfloxacin. They proposed that the cyclopropane ring that

ciprofloxacin possesses is able to alter the DNA gyrase activity, rendering it more

efficient [120].

Ciprofloxacin resistance is a highly common occurrence in S. aureus strains [91,

121, 122], as such it is not indicated for infections and the clinical ramifications of

SAGAR in S. aureus are currently benign. Quinolones, however, are the primary

treatment option for urinary tract infections caused by Escherichia coli [123]. Despite the

similarity in the quinolone resistance mechanism, this SAGAR phenotype has yet to be

reported in other bacteria besides S. aureus. In order to determine the mechanism

responsible for this SAGAR, and the significance of the structure of ciprofloxacin to the

phenotype, more quinolones/fluoroquinolones need to be tested. Studying other bacterial

species in which quinolones are still being used for therapeutic treatments can also

further elucidate the mechanism.

Ciprofloxacin is associated with DNA damage, specifically through double

stranded DNA breaks and stalled replication forks, which are processed to single-

stranded DNA [88]. Looking at the global transcriptional response to ciprofloxacin

treatment by Cirz et al. [88] revealed induction of Pol III and Y-family polymerases,

which emphasizes a common strategy of reduced metabolism and funneling of resources

into DNA synthesis in response to these antibiotics. It is also believed that DNA

synthesis might be error prone due to the down-regulation of mismatch repair genes

during exposure to ciprofloxacin [124, 125]. Therefore, it is possible that salicylate, when

coupled with ciprofloxacin, exacerbates this effect, which may explain the drug

specificity of SAGAR.

   30  

Through DNA sequencing of the ciprofloxacin targets, grlAB and gyrAB, our

results revealed mutations only in the QRDR domain of grlA. No difference in the level

of ciprofloxacin resistance was observed between LC and SCV colonies selected in the

presence or absence of salicylate. However, for norfloxacin, a higher level of resistance

was observed between LC selected in the absence of salicylate than those selected in the

presence of salicylate. This indicates a mutation outside of the grlA QRDR in the

salicylate treated colony that is accounting for this slightly decreased level of norfloxacin

resistance. Determining the exact mutation responsible for this difference is not practical

without resorting to whole genome sequencing [126-128].

S. aureus has been shown to result in heterogeneous expression of resistance to

various antibiotics [133, 57, and 134]. For example, heterogeneous intermediate

resistance to vancomycin (hVISA) is attributed to several genes related to cell regulation

pathways including vraSR, graSR saeSR, and agr [129-132]. A similar heterogeneity has

been observed by Price et al. [26] for salicylate inducible phenotypic resistance to fusidic

acid. The level of fusidic acid resistance induced by salicylate was found to be dependent

upon the genetic background of the strain.

It was thus no surprise when this study revealed that the SAGAR phenotype was

strain specific. Expression of SAGAR varied substantially between strains in this study,

as well as in previous studies for ciprofloxacin, ranging from a 8-fold to a 100-fold

increase in mutation frequency [33]. This variation likely reflects the inherent genetic

variation among these S. aureus strains and suggests that there are unknown genetic

factors important for this phenotype.

   31  

One of the genetic factors important for this heritable phenotype is likely to be σB,

as mutation of sigB significantly reduced expression of SAGAR for ciprofloxacin. σB

plays a prominent role in the cell, controlling roughly 198 genes [79], many of which are

involved in resistance to stressors [133]. σB has also been shown to be involved in the

repression of exoproteins and toxins, and is a positive regulator of adhesion factors [79,

80, 103, 133]. More importantly, σB is believed to play a role in mediating antibiotic

resistance [79]. Inactivation of sigB in MRSA-COL was found to increase its

susceptibility to methicillin [98] while mutations within the rsbU-defective strain BB255,

leading to SigB hyperproduction, were associated with an increase in glycopeptide

resistance [134]. σB has been shown under stress, to upregulate genes responsible for

maintaining cell integrity, membrane transport processes and intermediary metabolism

[118]. Thus, it is possible that salicylate in σB null strains is able to permeate the cell

more readily, allowing more salicylate to enter the cell, further enhancing its toxicity and

mitigating SAGAR.

S. aureus BB255, a ciprofloxacin sensitive strain carrying an 11-bp deletion in rsbU

encoding a positive regulator of σB [33, 135, 136], demonstrated a 100-fold increase in

mutation frequency to ciprofloxacin with salicylate [33]. SH1000 differs from BB255 in

that SH1000 is rsbU+, and has increased pigmentation, more vigorous growth, decreased

secreted exoproteins and decreased agr expression. Thus, perhaps differences in SAGAR

for ciprofloxacin between strains reflect differences in sigma B activity. However, the

frequency of mutation to ciprofloxacin resistance for SaRNH-1 (SH1000sigB::tet) was

actually less than 50% of that observed for SH1000, and less than 10% that of BB255.

Despite the 11-bp deletion in the rsbU gene, researchers have shown salicylic acid to

   32  

activate σB in both rsbU+ and rsbU null strains [137]. Therefore, it is possible that σB

positively influences SAGAR while rsbU negatively influences SAGAR. Our results

indicate σB dependence for full expression of the SAGAR phenotype.

The use of salicylate analogs to investigate the importance of chemical structure in

the phenotype revealed the importance of the ortho-hydroxyl group of salicylate. Sodium

benzoate lacks this hydroxyl group, while acetylsalicylic acid has an acetyl group in the

ortho position. Removal of the hydroxyl group mitigated SAGAR for both ciprofloxacin

and norfloxacin. While for lincomycin removal of the hydroxyl group (i.e. by use of the

structural analog benzoate) potentiated the effect salicylate had on reducing the

frequency at which resistance occurred. Interestingly, the reactive group of salicylate for

inducing oxidative stress in mitochondria has been shown to be this hydroxyl group

[138], which is believed to interact with a Fe-S cluster of mitochondrial Complex I

resulting in the production of ROS [14, 15, 82, 138]. A structural dependence was also

observed for salicylate-inducible phenotypic resistance to ciprofloxacin [33]. However,

the functional group was determined to be the carboxylic acid [33], which suggests a

potential mechanistic distinction between inducible phenotypic resistance and SAGAR.

The results suggest that this hydroxyl group is important for SAGAR, which could

possibly lead to ROS accumulation. This ROS accumulation could result in DNA

damage, which could explain the increase in mutation frequency.

Finally, our results indicate both inducible phenotypic resistance (data not shown)

and SAGAR require the presence of salicylate in the media for the phenotype, which

suggests a mechanistic link. It is possible that salicylate has a physiochemical effect on

the cell which leads to metabolic toxicity or that salicylate is a ligand for a yet to be

   33  

determined protein which can both affect antibiotic permeability results in interference

with DNA repair mechanisms.

   34  

CHAPTER THREE:

ROLE FOR METABOLIC STRESS IN THE SALICYLATE-ASSOCIATED

ANTIBIOTIC RESISTANCE PHENOTYPE

3.1 Background

Metabolic alterations play an important role in resistance of bacteria to

antimicrobials [139]. Changes in growth rate have been shown to influence bacterial

susceptibility to antibiotics [139]. For example, metabolically moribund E. coli have been

shown to be highly resistant to ampicillin or tetracycline, and many antibiotics have

reduced activity against stationary phase cultures [140]. Also, changes in bacterial

metabolism, such as those associated with dormancy or biofilm formation, are associated

with reduced susceptibility to antibiotics [141]. Alterations in metabolism in response to

antimicrobials have been also linked to small colony variant (SCV) formation in S.

aureus [142]. SCVs display a decreased rate of cell wall biosynthesis, a reduction in the

uptake of positively charged antimicrobials and an increase in survival in host cells [142].

These reductions are associated with a 4-fold increase in MICs to cell wall targeting

antibiotics [142].

Growth with salicylate induces phenotypic resistance to antimicrobials, and

increases the frequency at which genotypic resistance (SAGAR phenotype) to some

antibiotics occurs in S. aureus [7, 26, 33, 43, 70, 71]. These phenotypes occur at

   35  

concentrations of salicylate which are subinhibitory, but which are still toxic to growth.

Previous studies revealed substantial alterations in the transcript levels of genes

associated with central metabolism during growth with sub-MIC levels of salicylate [43].

Specifically, salicylate reduced the expression of genes important for

glycolysis/gluconeognesis such as gapA2 (encoding GAPDH) and pgi (glucose-6-

phosphate isomerase). In addition, salicylate increased expression of genes for gluconate

metabolism via the pentose phosphate shunt. This indicates that growth with salicylate at

concentrations that induce phenotypic resistance, and which are associated with SAGAR,

may inhibit glycolysis and increase gluconate metabolism [43]. Furthermore, these

findings suggest that S. aureus may alter flux through metabolic pathways to counter the

toxic effects of salicylate. In support of this claim, growth with gluconate was shown to

rescue the cell from the toxic effects of salicylate on growth [43]. Salicylate has been

shown to also result in ROS accumulation and DNA damage in eukaryotic models [5, 7,

82]. Specifically, salicylates have damaging effects on isolated mitochondria and have

been shown to result in uncoupling of oxidative phosphorylation as well as swelling

[138]. It is possible that salicylate has the same DNA damaging effects which can

ultimately lead to this increase in mutation frequency; i.e SAGAR. Collectively, these

studies reveal that salicylate may act at several levels to negatively impact metabolism in

the cell. The importance of these metabolic alterations in the presence of salicylate to

antibiotic resistance in S. aureus is currently unknown. The following experiments were

designed to examine the link between the metabolic toxicity of salicylate and the

SAGAR phenotype.

   36  

3.2 Methods

Generation Time Determination

The following protocol was adapted from Neoh et al.[132]. An overnight culture

of strain SH1000 was prepared under standard conditions before being subcultured into

fresh TSB to an optical density at 600 nm OD600=0.05. Optical density readings were

recorded from an OD600=0.05 to an OD600=0.5. Cells were treated with increasing

concentrations of NSAID, a weak acid, or for controls, an equal volume of water (or

respective solvent used for stocking NSAIDs/weak acids). Samples or plate counts

(CFU/ml) were taken every 30 min. Generation times were determined using Eq. 3.1

(below), where g is generation time, N1 is optical density at which the last reading was

taken, N0 is the initial optical density taken and t1 is the time in minutes from the initial

reading (t0)) to reach an optical density of 0.5. Data was analyzed using a t-test to identify

statistically significant differences between controls and treatment.

Eq. 3.1: g = !"# !"!#!!!!!

Mutation frequency determination

The following protocol was adapted from Foster, 2006 [108]. This protocol is

identical to that of Chapter 2 Methods, section 2.2, except that mutation frequency to the

antibiotic was determined over a range of NSAID and weak acid concentrations. For

mutation frequency determinations under anaerobic conditions, cultures were grown in

anaerobic chambers with CO2 packs (EZ container system, Becton, Dickson and

company) and incubated at 37°C.

   37  

Metabolite profiling

The following protocol was adapted from Meyer et al. 2010 [143]. Overnight

cultures (n=3) of S. aureus strain SH1000 were prepared under standard conditions. The

following day the culture was inoculated into 100 ml fresh TSB at OD600=0.05. This

culture was grown at 37°C at 448 x g until an OD600=0.5 was reached. At this point either

500 µg/ml or 2500 µg/ml of salicylate was added to the experimental flask or the

equivalent volume of water was added to the control flask and cultured for 30 min. From

each flask, 20 ml of culture was extracted and then filtered through a 0.22 µm Millipore

filter. Cells were then washed twice with 10 ml cold 0.6 % (wt/vol) NaCl. The filter was

then cut out and placed into a corning tube containing 10 ml of ice cold 60% (vol/vol)

ethanol. The corning tubes were vortexed for 10 minutes to ensure that all the cells came

off the filter. The filter was then discarded and the corning tubes were stored in the -80°C

freezer for an hour. Cells were then thawed on ice, while being rigorously mixed and

shaken 10 times alternately. Aliquots of 1 ml cell suspensions were transferred into an

appropriate number of tubes containing 0.5 cm3 glass beads (diameter 0.10-0.11 mm).

Cells were disrupted for 2 cycles for 30 seconds at 5,179 x g in a bead-mill (Mini-bead

beater, Biospec). After cell disruption the glass beads and the cell debris were separated

from the supernatant by centrifugation for 5 min at 4°C and 10,000 x g. The aliquoted

samples were combined, and the glass beads were washed once with 1 ml ddH2O each.

Washing entailed vortexing of the beads in the water and centrifugation for 5 min at 4°C

and 10,015 x g. The washing solutions were added to the combined samples. The

supernatant including the metabolites were brought to a final ethanol concentration of 10

% (vol/vol) and stored at -80°C. The metabolites were shipped on dry ice to the

   38  

University of Illinois for analysis. The metabolites were analyzed separated by liquid and

gas chromatography and analyzed by mass spectrometry at the Metabolomics Center,

Roy J. Carver Biotechnology Center, University of Illinois at Urbana-Champaign

(http://www.biotech.uiuc.edu/centers/MetabolomicsCenter/index.html).

Analysis of intracellular metabolites

Metabolite peaks were identified by comparison to a spectrum generated from

standards. The means and standard deviation were calculated for each treatment. An

ANOVA was performed comparing 0 µg/ml, 500 µg/ml and 2500 µg/ml salicylate

(a=0.05, n=3). A 2- fold change was used as a cutoff for biological significance. Those in

which statistics and biological significance were satisfied were further characterized

based on metabolic pathway. The online software Metaboanalyst was used to identify

pathways (http://www.metaboanalyst.ca/MetaboAnalyst/faces/Home.jsp).

Reactive oxygen species assay (ROS)

ROS was measured using luminol and DCFH-DA. Overnight cultures of S.

aureus strain SH1000 (n=3) were prepared under standard conditions. The following day

the culture was diluted into fresh TSB at a 1:100 to obtain OD600=0.05. The culture was

allowed to grow in a shaking incubator at 37°C at 448 x g to an OD=0.5. The appropriate

volumes of salicylate were added to obtain a final concentration of 500 and 2500 µg/ml.

These treatments served as the experimental conditions. Carbonyl cyanide m-

chlorophenylhydrazone (CCCP) was added to a final concentration of 100 µM, which

served as a positive control, and glycerol at 0.4% (vol/vol) as a negative control. An

   39  

additional treatment, which included only cells grown in TSB, served as a control for

basal ROS levels. Cells were exposed to each treatment for a total of 24 hours, and

readings were taken at 0.5 h, 3 h, 8 h, and 24 h. At each time point 1 ml of culture was

placed into a 1.5 ml microtube and centrifuged at 2,500 x g for 15 min. The supernatant

was decanted and the cells were resuspended in 500 µl of phosphate buffered saline

(PBS, pH 7.1).

For DCFH-DA, the following protocol was adapted from OxiSelect ROS Kit

(Cell Bio Labs, California) and performed per the manufacturer’s instructions. To each

sample, 2’,7’-dichlorfluorescein-diacetate (DCFH-DA) was added to a final

concentration of 10 µM. The samples were vortexed and incubated in the dark at 37°C

for 5 minutes. From each tube, 100 µl was added in triplicates to a 96 well plate. Luminol

was used for ROS detection following previously described methods (Chen et al. 2011).

Briefly, for each treatment, 100 µl of the prepared sample was added in triplicate to a 96

well plate. To each prepared well 100 µl of 250 µM luminol was then added. A synergy2

plate reader with Gen5 software (BioTek, Vermont) was used to record fluorescence.

Fluorescence was recorded at 460/40 emission and 360/40 excitation. The following

protocol was performed with a sample size of 3. Optical density for each treatment at

each time interval was recorded. Relative fluorescence units recorded were adjusted to

OD. Statistical significance between treatments was compared using an ANOVA

(a=0.05, n=3) (R ver 2.13.0).

   40  

NAD+/NADH Assay

To conduct a standard curve for NADH/NAD quantitation, 10 µl of a 1 nmol/µl

NADH standard (Biovision, California) was diluted with 990 µl NADH/NAD Extraction

Buffer to generate 10 pmol/µl standard NADH. Of the diluted NADH standard, 0, 2, 4, 6,

8, 10 µl were added into labeled 96-well plate in duplicates, resulting in a dilution series

of 0, 20, 40, 60, 80, 100 pmol/well. The final volume was adjusted to 50 µl with

NADH/NAD extraction buffer (Biovision). Readings were taken using a plate reader at

OD450. The standard curve was analyzed for strength of linearity by regression analysis

(R). For NADH/NAD quantitation, an overnight culture was prepared in TSB with S.

aureus strain SH1000. The following day 1 ml of cells was washed with 500 µl cold PBS

and pelleted at 448 x g for 5 min. To each sample, 400 µl of NADH/NAD extraction

buffer was added followed by two cycles of freeze/thaw (20 min at -80°C, then 10 min

room temperature). The cells and buffer were then vortexed for 10 sec and pelleted at

21,952 x g for 5 min. The extracted NADH/NAD supernatant was transferred into a

labeled tube. This was performed before addition of salicylate or CCCP, and 3 h

following addition of 1 µg ciprofloxacin, 500 µg/ml salicylate, 2500 µg/ml salicylate or

an equal volume of water (control).

To detect total NAD (NAD+ + NADH) 50 µl of the extracted samples were

transferred into a 96-well plate in duplicates. An NAD cycling mix was prepared for each

reaction by adding 100 µl of NAD cycling buffer and 2 µl of NAD cycling enzyme mix.

One hundred microliters of the NAD cycling mix was added to each well and incubated

at room temperature for 5 min. Ten microliters of NADH developer was added to each

well and incubated at room temperature for 1 hr. Readings were taken using a plate

   41  

reader at OD450. NADH was detected by adding 200 µl of the extracted samples into

Eppendorf tubes. The samples were then heated to 60°C for 30 min. This step

decomposed all NAD+. The samples were then cooled on ice and vortexed to remove

precipitates. Fifty microliters of the NAD+ decomposed samples were transferred into 96-

well plates in duplicates. One hundred microliters of the NAD cycling mix was added to

each well and incubated at room temperature for 5 min. Ten microliters of NADH

developer was added to each well and incubated at room temperature for 1 hr. Readings

were taken using a plate reader at OD450. Statistical significance was determined between

treatments using an ANOVA (a=0.05, n=3) (R ver 2.13.0). The sample readings were

applied to the NADH standard curve with the following equation X=(Y+0.0527)/0.3071.

The amount of NAD+ or NADH in the sample wells were calculated then divided by

OD600 of the culture prior to extraction.

Eq3.2: NAD/NADH Ration is calculates as:

3.3 Results

Dose-dependence of salicylate-associated genotypic antibiotic resistance

To determine the relationship between salicylate, sodium benzoate, and

acetylsalicylic acid growth toxicity and the salicylate associated genotypic antibiotic

resistance (SAGAR) phenotype, generation time (g) and mutation frequency (µ) were

calculated for a wide range of salicylate, sodium benzoate, and acetylsalicylic acid

concentrations. Salicylate was only shown to significantly impair growth at 2500 µg/ml,

NADtotal-NADH

NADH

   42  

increasing g from 29.7 to 48.8 min (p=0.002) (Fig. 3.1A). For Asa, this was slightly less

at 2000 µg/ml, increasing g to 54 min (p=0.002) (Fig. 3.1C), and benzoate was the least

growth toxic, requiring 5000 ug/ml to significantly inhibit growth, increasing g to 45.7

min (p=0.012) (Fig. 3.1E).

For Sal, Asa, and Ben, SAGAR was determined to be concentration-dependent,

and only occurred at non-growth-toxic concentrations. For Sal, SAGAR was observed at

50-1000 ug/ml, but was absent at 2500 ug/ml (Fig. 3.1B). Likewise, for Asa and Ben

SAGAR was observed at 50-500 µg/ml and 50-1000 ug/ml, but absent at 2000 µg/ml and

5000 µg/ml, respectively (Fig 3.1 D and F). Thus, the expression of the SAGAR

phenotype is suppressed at growth-toxic concentrations of Sal, Asa and Ben.

   43  

Figure 3.1. Dose dependency of salicylate, aspirin, and benzoate for increased frequency of resistance to ciprofloxacin. Generation time of SH1000 as a function of increasing concentrations of salicylate (panel A), aspirin (panel C) and benzoate (panel E). (B) Mutation frequency of SH1000 to ciprofloxacin resistance plotted against an increasing concentration of salicylate (panel B), aspirin (panel D) and benzoate (panel E). Asterisks denote statistical significance using a t-test (*p<0.05).

   44  

Metabolite profile of S. aureus grown in the presence of salicylate

To ascertain the metabolic alterations associated with salicylate exposure,

metabolite profiling was performed for S. aureus strain SH1000 exposed to nontoxic

(500 µg/ml) and growth-toxic (2500 µg/ml) concentrations of salicylate, which select for

genotypic resistance, or are non-selective, respectively. Metabolite profiling revealed a

total of 153 altered metabolites among the three treatments (Table 3.1). Of the 153

altered metabolites 88 were less abundant, 3 had no change, and 44 were more abundant

for the 2500 µg salicylate treatment compared to the control treatment. Also, 78

metabolites were less abundant, 3 had no change, while 61 were more abundant for the

500 µg/ml salicylate treatment compared to the control. Between the 2500 µg/ml and 500

µg/ml salicylate treatments 82 metabolites were less abundant, 4 had no change and 53

were more abundant for the 2500 µg/ml treatments when compared to the 500 µg/ml

treatment.

Metabolite profile revealed a decrease in TCA metabolites at the 500 µg/ml

salicylate treatment, as seen through a reduction in citric acid (0.4-fold), acontic acid

(0.2-fold), α-ketoglutaric acid, (0.4-fold) and fumaric acid (0.6-fold) in relation to non-

salicylate treated cultures, respectively (Table 3.1). A 2.1 fold accumulation of 2-

phosphoglycerate was also observed in the 500 µg/ml salicylate treatment when

compared to the 2500 µg treatment (p=0.04) (Table 3.1). Previously a study by Riordan

et al. 2007 [43] showed downregulation of gapA2 encoding a glyceraldehyde-3-

phosphate dehydrogenase, pgi encoding a glucose-6-phosphate isomerase, and

SACOL1838 encoding a phosphoenolpyruvate carboxykinase. Each is indicative of

reduction in glycolysis. Accumulation of 2-phosphoglycerates revealed further

   45  

downregulation of glycolysis as this gene is involved in the final step of converting

glucose to pyruvate. This is also corroborated in eukaryotic systems, specifically as seen

in mitochondria [84]. A 8.8-fold increase in 2-ketogluconic acid compared to non-

salicylate treated cultures was observed at the 500 µg/ml treatment as well (p=0.023)

(Table 3.1). The S. aureus gluconate operon gntRKP was shown to be upregulated with

salicylate stress [43]. This increase in gluconate utilization gene expression may reflect

an important metabolic alteration necessary to compensate for the growth inhibitory

effects of salicylate. This effect on gluconate metabolism genes was also observed at the

2500 µg/ml salicylate treatment (p=0.00019), with a 6.7-fold increase in levels compared

to the non-salicylate treated treatment (Table 3.1). Interestingly, an accumulation of lactic

acid was observed with salicylate treated cultures, which is indicative of anaerobic

fermentation (p=0.035) [145]. In addition, butanediol increased by 4-fold in the presence

of salicylate (Table 3.1). Lactic acid and butanediol production are used by S. aureus to

cope with acid-stress [146], suggesting that perhaps growth with salicylate leads to

acidification of the cytoplasm. For example, to avoid further acidification of an already

acidic internal environment pyruvate is metabolized via acetolactate or the diacetyl

pathway to butanediol. It is believed that S. aureus increases pH through accumulation of

ammonium and through the removal of acid groups, which results in the production of

2,3-butanediol [146]. Also, the accumulation of acid may aid in stress as lactic acid was

initially synthesized from pyruvate, which can then be oxidized further by the TCA cycle

[141].

Two important mechanisms for increasing pH in S. aureus cultures which are acid

stressed is through the production of ammonia by urease and removal of acids [146]. An

   46  

increase in urea was observed in the presence of 500 µg/ml salicylate by 1.8-fold

(P=0.0026) (Table 3.1). A dose dependent effect was observed between the 500 and 2500

µg/ml treatments. Specifically, a 5-fold increase in uracil was observed at the 500 µg/ml

treatment when compared to the 2500 µg/ml treatment indicating acid-stress. Also, 2-

phosphoglycerate was observed to be 2-fold higher at 500 µg/ml than 2500 µg/ml

indicating impairment in glycolysis. Butanediol was 4-fold higher at the 2500 µg/ml

treatment than the 500 µg/ml treatment again indicating acid stress.

   47  

Table 3.1. Metabolite profile for S. aureus grown with salicylate

Metabolite

Salicylate concentration

Confidence Interval

(µg/ml) (CI) Relative Abundance

1,2-Benzenedicarboxylic acid 2500 500 2500 500 1-Ethylglucopyranoside 2.3 2.6 1.11 0.39 1-Methyl-beta-D-galactopyranoside NA 1.1 NA NA 1-Monooctadecanoylglycerol 0.4 0.7 NA 0.16 2(1H)-Pyrimidinone, 1-D-ribofuranosyl-4-hydroxy-5'-p 1.4 1 0.38 0.38

2,3-hydroxybutane 0.4 0.5 0.98 0.38 2,3-hydroxysuccinic acid 1.2 1.3 0.15 0.1 2,4,6-hydroxypyrimidine 1.3 1.1 0.33 0.33 2,4,6-Tri-tert.-butylbenzenethiol 0.5 1.5 0.38 0.22 2,4-hydroxybenzoic acid 1.7 1.7 0.21 0.69 2,4-hydroxybutanoic acid 3.3 2.7 0.65 0.98 2-Amino-4,6-dihydroxypyrimidine 1 0.9 0.65 0.65 2-Aminobutyric acid 1.2 1.7 0.15 0.35 2-aminoethylphosphoglycerate NA 0.5 0.37 0.68 2-Furancarboxylic acid 0.5 NA NA 0.15 2-hydroxybenzoic acid 0.7 0.9 0.34 NA 2-hydroxybutanoic acid 528.2 166.6 0.16 0.16 2-Hydroxyglutaric acid 0.9 1 306.07 176.77 2-Ketogluconic acid 0.6 0.8 0.31 0.12 2-methyl-2,3-hydroxypropanoic acid 6.7 8.8 0.12 0.29 2-methyl-2-hydroxybutanoic acid 0.7 0.7 10.43 7.36 2-methylbenzoic acid 1.3 1.7 0.03 0.24 2-oxo-3-hydroxypropanoic acid 1.9 1.9 0.95 1.52 2-oxophosphoglycerate 0.8 0.6 0.38 0.39 2-phosphoglycerate 0.9 1.1 0.57 0.36 3,4,5-Trihydroxypentanoic acid 0.1 0.3 0.32 0.12 3,4-Dihydroxybutanoic acid 2.4 4.7 0.03 0.11 3-methyl-3-hydroxybutanoic acid 1.2 2.1 0.2 2.72 3-phosphoglycerate 1.3 1.4 0.14 1.32 4,5-dimethyl-2,6-hydroxypyrimidine 1 1.1 0.2 0.21 4-Hydroxybutanoic acid 0.4 0.2 0.3 0.46 aconitic acid 0.9 1.3 0.29 0.15

   48  

Table 3.1 continued

Metabolite Salicylate concentration Confidence Interval

(µg/ml) (CI) Relative Abundance

Adenosine NA 0.3 0.13 0.12 Adenosine-5-monophosphate 0.2 0.6 NA 0.07 a-Glycerophosphate 0.6 0.8 0.12 0.09 Agmatine 0.4 1.1 0.14 0.16 a-ketoglutaric acid 0.5 0.4 0.1 0.1 Alanine 0.5 0.4 0.09 0.11 Aminomalonic acid 0.8 0.5 0.15 0.11 arabitol 0.1 0.4 0.19 0.07 Asparagine 0.9 1.2 0.18 0.21 Aspartic acid 0.3 0.3 0.04 0.28 B-alanine 1.1 1.2 0.11 0.14 Benzoic acid 1.3 1.2 0.2 0.21 Butylamine 1.3 1.4 0.45 0.07 citric acid 1.1 1 0.16 0.21 Diethyleneglycol 0.4 0.4 0.33 0.15 digalactosylglycerol 1.1 1.3 0.08 0.11 Eicosanoic acid 0.6 1 NA NA erythronic acid 1.6 1.3 0.26 0.39 ethanolamine 1 1.3 0.09 0.48 Ethyl phosphoric acid 0.9 1 NA NA Ferulic acid 0.4 0.8 0.76 0.45 fructose 1 0.7 0.14 0.44 Fructose-6-phosphate 0.2 0.3 0.25 0.55 Fumaric acid 1 2.3 0.09 0.51 Galactaric acid 0.7 0.6 0.23 0.04 galactose 0.3 0.2 0.13 0.21 Glucaric acid 0.5 0.7 0.24 1.28 Glucoheptulose 0.2 0.3 0.1 0.15 Gluconic acid 0.5 0.9 0.13 0.09 glucose 1.6 1.6 0.18 0.11 glucose-6-phosphate 0.2 0.5 0.07 0.17 Glutamic acid 0.4 0.3 0.25 0.65 Glutaric acid 0.3 0.5 0.79 0.67 galactose 0.3 0.2 0.13 0.21 Glucaric acid 0.5 0.7 0.24 1.28 Glucoheptulose 0.2 0.3 0.1 0.15

   49  

Table 3.1 continued

Metabolite Salicylate concentration Confidence Interval

(mg/ml) (CI) Relative Abundance

Gluconic acid 0.5 0.9 0.13 0.09 glucose 1.6 1.6 0.18 0.11 glucose-6-phosphate 0.2 0.5 0.07 0.17 Glutamic acid 0.4 0.3 0.25 0.65 Glutaric acid 0.3 0.5 0.79 0.67 Glyceric acid 0.8 1 0.11 0.34 Glycerol 0.9 0.8 0.22 0.11 Glycine 1 2.3 0.16 0.36 glycolic acid 0.5 0.5 NA NA Glycopyranose 0.9 1.1 0.04 0.23 Guanine 0 0.2 0.32 0.22 Hexadecanoic acid 1.5 0.9 0.93 1.75 Hexanoic acid 1.2 1 0.47 0.1 Hydroxylamine 0.2 0.7 0.69 0.26 Hydroxyphosphinyloxy-acetic acid 0.7 1.1 0.87 0.36 Hydroxyproline 1.2 1.1 0.69 0.53 Inositol, chiro- 0.8 1.4 0.29 0.33 inositol, myo- 0.7 0.8 0.06 0.25 Inositol, scyllo- 0.6 0.7 0.11 0.44 isoleucine NA 1 NA NA Isoxanthopterin 0.7 0.7 0.3 1 lactic acid 3.2 0.9 0.13 0.16 lactose 1.7 1.1 0.31 0.18 leucine 0.7 0.3 NA 0.68 lysine 0.6 1.2 0.09 0.3 Maleic acid 0.6 0.3 3.4 0.16 Malic acid 1.1 1.2 0.7 0.05 Malonic acid 0.4 0.6 0.25 0.14 Mannitol 1 0.8 0.19 1.47 mannose 1.7 1.8 0.29 0.2 mannose-6-phosphate 2 1.1 0.33 0.37 Melibiose 0.9 0.9 0.17 0.49 methionine 1 1.2 0.15 0.23 Methylmalonic acid 0.6 0.7 0.55 1.18 Monomethylphosphate NA 1.1 1.34 0.05 N-Acetyl aspartic acid 1.2 1 0.38 0.45 N-Acetylglutamic acid 0.7 1.1 0.62 0.82

   50  

Table 3.1 continued

Metabolite Salicylate concentration Confidence Interval

(mg/ml) (CI) Relative Abundance

N-Acetyl-Lysine 0.6 0.9 0.4 0.16 N-Acetyl-serine 0.7 0.6 NA 0.54 Nicotinic acid 1 1 0.3 0.54 Nonanoic acid 0.9 0.8 0.5 0.13 Octadecanoic acid 1.1 0.9 0.33 0.04 Octadecanol 1.5 1.3 0.21 0.2 ornitine 0.8 0.9 0.32 0.11 Orotic acid 0.2 0.5 0.46 0.09 oxalic acid 1 0.8 0.63 0.57 Panthotenic acid 0.5 0.5 0.99 0.45 phenylalanine 0.6 0.6 0.11 0.11 phosphoric acid 0.6 0.7 0 0.54 Pinitol 0.6 0.6 NA 0.22 Pipecolic acid 0.8 1.2 0.56 0.37 Pyroglutamic acid 1.2 1.9 0.24 0.09 pyrophosphate 0.6 0.5 0.12 0.28 Pyrrole-2-carboxylic acid 3 1.9 NA NA pyruvic acid 0.5 1.3 0.13 0.32 ribitol 0.7 1.1 0.48 0.04 ribose 0.8 1 0.28 0.39 ribose-5-p 1.3 2.3 0.2 1.13 Sedoheptulose 0.9 1 0.15 0.2 serine 0.8 0.8 2.26 0.59 Sorbitol 0.6 0.6 NA 0.33 sorbose 0.5 0.8 0.15 0.21 Succinic acid 0.6 1 0.6 0.52 sucrose 1 1 0.51 1.7 Threonic acid 0.3 0.2 0.17 0.41 Threonine 4.4 5.2 0.1 0.06 Thymine 0.9 1.1 0.35 0.25 Trehalose 0.5 0.5 0.55 0.11 Tryptophan 0.5 0.2 0.18 0.12 tyrosine 0.5 0.3 0.32 0.43 Uracil 0.7 0.5 0.09 0.16

   51  

Table 3.1 continued

Metabolite Salicylate concentration Confidence Interval

(mg/ml) (CI) Relative Abundance

Uridine 1.2 0.7 NA NA valine 0.4 0.8 NA NA vanillic acid 0.7 0.8 6.49 3.17 xylitol 1 1 0.14 0.7 All values mean fold change and are relative to the control, non-salicylate treated treatment. Not recorded (NA).

The effects of salicylate on the TCA cycle

Metabolite profiling revealed down regulation of glycolysis as well as the TCA

cycle. We therefore wanted to determine the effect of salicylate on the TCA cycle. These

findings correlated to a significant increase in levels of NAD+ in cells treated with 2500

µg/ml salicylate for 3 hours with a p=0.036 (Fig. 3.2). Ciprofloxacin, a positive control,

has been shown to result in an increase in NAD+ levels [89]. No increase in NAD+ levels

was detected at the 500 µg/ml treatment of salicylate. Oxygen dependence has been

linked to downregulation of the electron transport chain and the TCA genes and can be

seen through a decrease in NADH levels (Fig. 3.3) as well as the buildup of lactic acid

and butanediol as we have seen through the metabolite profile (Table 3.1) [145].

   52  

Figure 3.2. Effect of salicylate on cellular NAD+ levels. Mean (n≥3) percent nicotinomide adeninine dinucleotide (NAD+) plotted for ciprofloxacin and different salicylate concentrations, at time 0 (filled) and 3 h after treatment (stippled). Asterisks denotes statistical significance p<0.05 when compared to identical treatment at time 0. Oxygen dependence on the SAGAR phenotype

Oxygen has been shown to play an important role in the persistence and overall

growth of S. aureus in different conditions [145]. We were interested in determining the

oxygen dependence on the genotypic heritable phenotype. As shown in Fig. 3.3, in the

presence of oxygen there were 6.5-fold more cells recovered on 1 µg cipro plus 500

µg/ml salicylate plates when compared to the number of cells recovered in the absence of

oxygen, 2.6 x 10-6 to 3.9 x 10-7. This shows an inverted phenotype in the absence of

oxygen, which implies a dependence on oxygen or aerobic growth for this phenotype to

   53  

be expressed. On the other hand, at 2500 µg/ml of salicylate and 1 ug ciprofloxacin, there

was no difference observed between treatments. This indicates that the effect observed at

2500 µg/ml of salicylate is linked to toxicity rather than oxygen availability (Fig 3.3).

Also, the SAGAR phenotype was recapitulated in that the addition of 500 µg/ml

salicylate increased frequency of ciprofloxacin resistant isolates by 10.3-fold when

compared to without salicylate aerobically.

Figure 3.3. Expression of SAGAR during anaerobic growth. Mean (n=3) fold change in the frequency of mutation to CipR for anaerobic cultures relative to aerobic cultures plotted for growth on ciprofloxacin (control), and ciprofloxacin with 500 µg/ml or 2500 µg/ml salicylate. Role for reactive oxygen in the SAGAR phenotype

Salicylate has been shown to result in reactive oxygen species (ROS)

accumulation in mitochondria [82, 147]. ROS damages DNA resulting in oxidation of

   54  

guanine to 8-oxo-7, 8-dihydro-guanine, which can result in mutations [148]. To gain

further insight into the role for ROS in the SAGAR phenotype, ROS levels were recorded

using either DCFH-DA (Fig. 3.4 A) or luminol (Fig. 3.4 B). ROS levels from salicylate

stressed cells were compared to levels recorded from 100 µM CCCP, a positive control,

and the control treatment, which were SH1000 cells grown in the absence of any stressor.

A statistical significance with a p<0.05 was only observed with the positive controls (Fig.

3.4 AB). Specifically, a 2.7-fold and 6-fold increase in ROS was observed for CCCP

when compared to the control for DCFH-DA and luminol, respectively. However, no

increase in ROS was observed for both DCFH-DA and luminol for 500 and 2500 µg/ml

salicylate or the negative control, 04 % glycerol.

Fluoroquinolones are known to result in ROS accumulation [89]. We

hypothesized that since the genotypic heritable phenotype is only observed in the

presence of salicylate and an antibiotic such as ciprofloxacin, that we would detect ROS

in the presence of both drugs. Therefore, combinational effects of salicylate with

ciprofloxacin were tested for ROS levels (Fig. 3.4 C). ROS accumulation was observed at

1 µg/ml ciprofloxacin and also when combined with salicylate. However, ROS levels

were lower in the presence of salicylate when combined with ciprofloxacin than with

ciprofloxacin alone.

Recently, in a study by Paez et al. 2010 [92], the addition of glutathione to

ciprofloxacin treated cells was able to significantly reduce the MIC to ciprofloxacin

resistant S. aureus cells. The ability of salicylate to induce ROS was addressed with the

addition of glutathione, an antioxidant. The addition of glutathione to ciprofloxacin and

salicylate stressed cells, as expected, mitigated the number of CFUs/ml recovered (Fig.

   55  

3.5) indicating antioxidant properties for glutathione. However, the addition of

glutathione to ciprofloxacin alone or ciprofloxacin and salicylate treatments did not

decrease ROS levels (Fig. 3.4 C) indicating that glutathione was mitigating the SAGAR

phenotype in a different manner. Therefore, we believe that glutathione is not reducing

ROS, but rather altering the structure of salicylate, which impairs the ability of salicylate

to induce its genotypic effects.

   56  

 Figure 3.4. Salicylate associated ROS accumulation. ROS accumulation using: (A) DCFH-DA (B) luminol. (C) Combinational effects of ciprofloxacin and salicylate on accumulation of ROS using luminol. Asterisks denote statistical significance compared to control with a p<0.05, n=3. All values were recorded after 3 hours of stress under each treatment. GSH indicates 10 mM glutathione. Cipro indicates 1 µg/ml ciprofloxacin. Salicylate indicates 500 µg/ml salicylate. CCCP indicates carbonyl cyanide m-chlorophenylhydrazone.

   57  

Figure 3.5. The effects of glutathione on the SAGAR phenotype. Control treatment was strain SH1000 grown strictly in TSB. SH1000 cells were plated on TSA plates containing the above treatments. CFUs recovered were recorded. Asterisks denotes statistical significance with P<0.05 between cipro and salicylate treatment compared to cipro, salicylate and glutathione. GSH indicates 10 mM glutathione. Cipro indicates 1 µg/ml ciprofloxacin. Salicylate indicates 500 µg/ml salicylate.

3.4 Discussion

The results of this study revealed that the SAGAR phenotype is sensitive to

salicylate concentration; the phenotype is expressed at non-toxic concentrations, but is

suppressed at growth-toxic concentrations. Growth toxic concentrations were associated

with a metabolic switch to anaerobiosis, supported by the accumulation of lactic acid and

butanediol in our metabolite profile. Accumulation of such metabolites implies that this

   58  

concentration of salicylate induced weak acid stress [138]. Mixed acid (lactate, formate,

and acetate) and butanediol fermentation in S. aureus occur under anaerobic

conditions[138]. Pyruvate from glycolysis can be reduced to either lactate by activity of

lactate dehydrogenase or metabolized to acetoin and 2,3-butanediol by the activity of

acetolactate synthase (BudB), α-acetolactate decarboxylase (BudA1), and acetoin

reductase (SACOL0111) [138]. This process requires the oxidation NADH, which is a

requisite under fermentation conditions [138]. 2,3-butanediol is involved in a variety of

physiological activities such as homeostasis of pH and regulation of cellular

NAD/NADH ratio in bacteria [145].

The anaerobic effect induced by growth toxic concentrations of salicylate led to

question if the impairment in generation time was a causative or correlative effect. Under

anaerobic conditions SAGAR for ciprofloxacin was substantially reduced suggesting a

requirement of oxygen for the phenotype. Oxygen is necessary for the formation of a

functional electron transport system [150]. It is possible that the lack of oxygen prevents

ROS as well as DNA damage from occurring, i.e. an elevated mutation frequency, since

regulation of the electron transport chain is vital to the homeostasis of S. aureus. This

reveals that SAGAR requires an oxidative environment.

ROS damages iron-sulfur clusters making ferrous iron available for oxidation by

the Fenton reaction [84, 85]. The Fenton reaction leads to hydroxyl radical formation.

The hydroxyl radicals damage DNA, proteins, and lipids, which results in cell death [84,

85]. In a study by Chatterjee et al. inactivation of fur, a ferric uptake regulator homolog,

decreased butA, which decreased 2,3-butanediol productions, as well as the TCA cycle

   59  

genes citC, isocitrase [147]. These findings support the idea that toxic concentrations of

salicylate inhibit the TCA cycle, therefore, impairing the reduction of NAD+.

Salicylate in mitochondria has been shown to interact with the respiratory chain

resulting in hydrogen peroxide and other ROS which in turn oxidize thiol groups and

glutathione [131]. This oxidative stress leads to the induction of the mitochondrial

permeability transition in the presence of Ca2+. This leads to further increase of oxidative

damage, resulting in impairment of oxidative phosphorylation [131]. Based on these

findings we hypothesized that salicylate would result in ROS accumulation in S. aureus

cells. However, contrary to our hypothesis, induction of growth toxic and non-growth

toxic concentrations of salicylate did not result in detection of ROS using either

chemiluminescent DCFH-DA or luminol. Fluoroquinolones, specifically ciprofloxacin,

have been shown to result in ROS production, which is one of their main bactericidal

characteristics [85, 105, 106, 108, 146, 148, 149]. Considering that the mutation

frequency observed is always in the presence of an antibiotic, such as ciprofloxacin, we

hypothesized that the combination of salicylate and ciprofloxacin would result in a

significant increase in ROS. We however, did not observe ROS accumulation upon the

combination of both drugs. Also, the addition of glutathione, an antioxidant, did not alter

the levels of ROS produced for both ciprofloxacin and salicylate, which leads to conclude

that salicylate does not result in detectable ROS accumulation in S. aureus. However, the

addition of glutathione to ciprofloxacin and salicylate treatments did substantially

mitigate the SAGAR phenotype. This suggests that glutathione may not be acting like an

antioxidant, but rather could be altering the structure of salicylate. Alteration in the

   60  

structure of salicylate, as seen in Chapter 2 results section 2.3, also mitigates the SAGAR

phenotype.

A significant increase in mutation frequency is observed at non-growth toxic

concentrations of salicylate (500 µg). Despite the lack of ROS accumulation, metabolite

profiling revealed a decrease in glycolysis and TCA cycle as seen in decreased levels of

citric acid, acontic acid, α-ketoglutaric acid, and fumaric acid, which was not mirrored in

the increase in NAD+ levels at 500 µg/ml of salicylate. However, at growth toxic

concentrations of salicylate (2500 µg/ml), a significant increase in NAD+ levels was

observed, indicating either an increase in the oxidation of NADH or rather impairment in

TCA cycle, which is responsible for reducing NAD+. It is possible that the disruption of

the TCA cycle in addition to the growth impairment effects of salicylate at 2500 µg/ml in

combination are responsible for the toxicity and in result are responsible for mitigation of

the SAGAR phenotype.

Another effect observed with toxic concentrations of salicylate by Riordan et al.

[42] is a significant decrease in transcription of glycolytic genes gapA2 and pgi, which

are important genes in this process. Also, a significant increase in the gluconate operon

(gntkPR) was observed upon salicylate stress [42]. Interestingly, Riordan et al. [42]

observed exacerbation and reduction in growth inhibitory effects of salicylate upon

glucose and gluconate addition, respectively. We cared to further explore this observation

with a wide range of fermentable and non-fermentable sugar sources in a sugar free

media, CASY (data not shown). We did not see any sugar that significantly impaired

growth toxic effects of salicylate when compared to treatments of salicylate in the

absence of a sugar source. We did however; observe that glucose exacerbated the

   61  

inhibitory effects of growth toxic concentrations of salicylate (data not shown). We

hypothesized that this effect was due to a decrease in pH, which altered membrane

permeability and in result enhancing the toxicity of salicylate. To test this hypothesis we

used a buffer, MOPS to determine if it was in fact an acid effect (data not shown). As

hypothesized, lowering the pH decreased the concentration of salicylate needed to inhibit

growth. This finding leads us to believe that it is in fact the acidification of media that

results in alteration of membrane permeability. This finding was observed in Serratia

marcescens [89]. The effect of salicylate has been attributed to a weak acid effect that

possibly leads to an increase in the membrane potential [89].

   62  

REFERENCES 1.   Alvarez,   L.P.,   et   al.,   Salicylic   acid   diminishes   Staphylococcus   aureus   capsular  polysaccharide  type  5  expression.  Infect  Immun,  2010.  78(3):  p.  1339-­‐44.  

 2.   Bennett,  A.  and  I.A.  Tavares,  NSAIDs,  Cox-­‐2  inhibitors,  and  the  gut.  Lancet,  1995.  346(8982):  p.  1105.  

 3.   Bennett,   J.S.,   et   al.,  The  use  of  nonsteroidal  anti-­‐inflammatory  drugs   (NSAIDs):  a  science  advisory  from  the  American  Heart  Association.  Circulation,  2005.  111(13):  p.  1713-­‐6.  

 4.  Mazumdar,   K.,   et   al.,   Diclofenac   in   the   management   of   E.   coli   urinary   tract  infections.  In  Vivo,  2006.  20(5):  p.  613-­‐9.  

 5.   Smith,  C.C.,  et  al.,  NSAIDs  and  gut  toxicity.  Lancet,  1994.  344(8914):  p.  56-­‐7.    6.  Mahmud,  T.,  D.L.  Scott,  and  I.  Bjarnason,  A  unifying  hypothesis  for  the  mechanism  of  NSAID  related  gastrointestinal  toxicity.  Ann  Rheum  Dis,  1996.  55(4):  p.  211-­‐3.  

 7.   Price,  C.T.,   I.R.   Lee,   and   J.E.  Gustafson,  The  effects  of  salicylate  on  bacteria.   Int   J  Biochem  Cell  Biol,  2000.  32(10):  p.  1029-­‐43.  

 8.   Perry,  C.A.,  et  al.,  Health  effects  of  salicylates  in  foods  and  drugs.  Nutr  Rev,  1996.  54(8):  p.  225-­‐40.  

 9.   Pearlman,  B.L.  and  R.  Gambhir,  Salicylate  intoxication:  a  clinical  review.  Postgrad  Med,  2009.  121(4):  p.  162-­‐8.  

 10.  Patel,  D.K.,  et  al.,  Metabolism  of  aspirin  after  therapeutic  and  toxic  doses.  Hum  Exp  

Toxicol,  1990.  9(3):  p.  131-­‐6.    11.  Oh,   K.W.,   et   al.,   Salicylate   enhances   necrosis   and   apoptosis   mediated   by   the  

mitochondrial  permeability  transition.  Toxicol  Sci,  2003.  73(1):  p.  44-­‐52.    12.  Shiff,  S.J.  and  B.  Rigas,  Aspirin  for  cancer.  Nat  Med,  1999.  5(12):  p.  1348-­‐9.    13.  Al-­‐Nasser,   I.A.,  Salicylate-­‐induced  kidney  mitochondrial  permeability  transition  is  

prevented  by  cyclosporin  A.  Toxicol  Lett,  1999.  105(1):  p.  1-­‐8.  

   63  

14.  Trost,   L.C.   and   J.J.   Lemasters,  The  mitochondrial   permeability   transition:   a   new  pathophysiological   mechanism   for   Reye's   syndrome   and   toxic   liver   injury.   J  Pharmacol  Exp  Ther,  1996.  278(3):  p.  1000-­‐5.  

 15.  Isenberg,   J.S.   and   J.E.  Klaunig,  Role  of  the  mitochondrial  membrane  permeability  

transition   (MPT)   in   rotenone-­‐induced   apoptosis   in   liver   cells.   Toxicol   Sci,   2000.  53(2):  p.  340-­‐51.  

 16.  Bernardi,  P.,  Modulation  of  the  mitochondrial  cyclosporin  A-­‐sensitive  permeability  

transition  pore  by  the  proton  electrochemical  gradient.  Evidence  that  the  pore  can  be  opened  by  membrane  depolarization.  J  Biol  Chem,  1992.  267(13):  p.  8834-­‐9.  

 17.  Bernardi,  P.,  K.M.  Broekemeier,  and  D.R.  Pfeiffer,  Recent  progress  on  regulation  of  

the  mitochondrial  permeability  transition  pore;  a  cyclosporin-­‐sensitive  pore  in  the  inner  mitochondrial  membrane.  J  Bioenerg  Biomembr,  1994.  26(5):  p.  509-­‐17.  

 18.  Bernardi,   P.,   P.   Veronese,   and   V.   Petronilli,   Modulation   of   the   mitochondrial  

cyclosporin  A-­‐sensitive  permeability   transition  pore.   I.   Evidence   for   two   separate  Me2+  binding  sites  with  opposing  effects  on  the  pore  open  probability.  J  Biol  Chem,  1993.  268(2):  p.  1005-­‐10.  

 19.  Kroemer,   G.,   et   al.,  The  biochemistry   of   programmed   cell   death.   FASEB   J,   1995.  

9(13):  p.  1277-­‐87.    20.  Pugliese,   A.,   T.   Beltramo,   and   D.   Torre,   Reye's   and   Reye's-­‐like   syndromes.   Cell  

Biochem  Funct,  2008.  26(7):  p.  741-­‐6.    21.  Glasgow,   J.F.,  Reye's   syndrome:   the   case   for  a   causal   link  with  aspirin.   Drug   Saf,  

2006.  29(12):  p.  1111-­‐21.    22.  Hall,   S.M.,   et   al.,   Preadmission   antipyretics   in   Reye's   syndrome.   Arch   Dis   Child,  

1988.  63(7):  p.  857-­‐66.    23.  McGovern,   M.C.,   J.F.   Glasgow,   and   M.C.   Stewart,   Lesson   of   the   week:   Reye's  

syndrome  and  aspirin:  lest  we  forget.  BMJ,  2001.  322(7302):  p.  1591-­‐2.    24.  Thabet,  F.,  et  al.,  [Severe  Reye  syndrome:  report  of  14  cases  managed  in  a  pediatric  

intensive  care  unit  over  11  years].  Arch  Pediatr,  2002.  9(6):  p.  581-­‐6.    25.  Cohen,   S.P.,   et   al.,  Salicylate  induction  of  antibiotic  resistance  in  Escherichia  coli:  

activation  of  the  mar  operon  and  a  mar-­‐independent  pathway.   J  Bacteriol,  1993.  175(24):  p.  7856-­‐62.  

 26.  Price,  C.T.,  et  al.,  Effects  of  salicylate  and  related  compounds  on  fusidic  acid  MICs  in  

Staphylococcus  aureus.  J  Antimicrob  Chemother,  1999.  44(1):  p.  57-­‐64.  

   64  

27.  Sumita,   Y.   and   M.   Fukasawa,   Transient   carbapenem   resistance   induced   by  salicylate   in   Pseudomonas   aeruginosa   associated   with   suppression   of   outer  membrane  protein  D2  synthesis.  Antimicrob  Agents  Chemother,  1993.  37(12):  p.  2743-­‐6.  

 28.  Sulavik,  M.C.,  M.  Dazer,   and  P.F.  Miller,  The  Salmonella  typhimurium  mar   locus:  

molecular  and  genetic  analyses  and  assessment  of  its  role  in  virulence.  J  Bacteriol,  1997.  179(6):  p.  1857-­‐66.  

 29.  Rosner,   J.L.,   Nonheritable   resistance   to   chloramphenicol   and   other   antibiotics  

induced   by   salicylates   and   other   chemotactic   repellents   in   Escherichia   coli   K-­‐12.  Proc  Natl  Acad  Sci  U  S  A,  1985.  82(24):  p.  8771-­‐4.  

 30.  Masuda,  N.,   E.   Sakagawa,   and  S.  Ohya,  Outer  membrane  proteins  responsible   for  

multiple   drug   resistance   in   Pseudomonas   aeruginosa.   Antimicrob   Agents  Chemother,  1995.  39(3):  p.  645-­‐9.  

 31.  Ma,   D.,   et   al.,   Genes   acrA   and   acrB   encode   a   stress-­‐induced   efflux   system   of  

Escherichia  coli.  Mol  Microbiol,  1995.  16(1):  p.  45-­‐55.    32.  Hutsul,   J.A.   and   E.   Worobec,   Molecular   characterization   of   the   Serratia  

marcescens   OmpF   porin,   and   analysis   of   S.   marcescens   OmpF   and   OmpC  osmoregulation.  Microbiology,  1997.  143  (  Pt  8):  p.  2797-­‐806.  

 33.  Gustafson,   J.E.,   et   al.,   Growth   in   the   presence   of   salicylate   increases  

fluoroquinolone   resistance   in   Staphylococcus   aureus.   Antimicrob   Agents  Chemother,  1999.  43(4):  p.  990-­‐2.  

 34.  Fralick,  J.A.,  Evidence  that  TolC  is  required  for  functioning  of  the  Mar/AcrAB  efflux  

pump  of  Escherichia  coli.  J  Bacteriol,  1996.  178(19):  p.  5803-­‐5.    35.  Domenico,   P.,   T.   Hopkins,   and   B.A.   Cunha,   The   effect   of   sodium   salicylate   on  

antibiotic   susceptibility   and   synergy   in   Klebsiella   pneumoniae.   J   Antimicrob  Chemother,  1990.  26(3):  p.  343-­‐51.  

 36.  Cohen,   S.P.,   et   al.,   Cross-­‐resistance   to   fluoroquinolones   in   multiple-­‐antibiotic-­‐

resistant   (Mar)   Escherichia   coli   selected   by   tetracycline   or   chloramphenicol:  decreased   drug   accumulation   associated  with  membrane   changes   in   addition   to  OmpF  reduction.  Antimicrob  Agents  Chemother,  1989.  33(8):  p.  1318-­‐25.  

 37.  Burns,   J.L.   and   D.K.   Clark,   Salicylate-­‐inducible   antibiotic   resistance   in  

Pseudomonas  cepacia  associated  with  absence  of  a  pore-­‐forming  outer  membrane  protein.  Antimicrob  Agents  Chemother,  1992.  36(10):  p.  2280-­‐5.  

   65  

38.  Repaske,  D.R.  and  J.  Adler,  Change  in  intracellular  pH  of  Escherichia  coli  mediates  the  chemotactic  response  to  certain  attractants  and  repellents.   J  Bacteriol,  1981.  145(3):  p.  1196-­‐208.  

 39.  Foulds,   J.,   et   al.,   Decreased   permeation   of   cephalosporins   through   the   outer  

membrane  of  Escherichia  coli  grown  in  salicylates.  Antimicrob  Agents  Chemother,  1989.  33(4):  p.  412-­‐7.  

 40.  Pomposiello,   P.J.,   M.H.   Bennik,   and   B.   Demple,   Genome-­‐wide   transcriptional  

profiling   of   the   Escherichia   coli   responses   to   superoxide   stress   and   sodium  salicylate.  J  Bacteriol,  2001.  183(13):  p.  3890-­‐902.  

 41.  Duy,   N.V.,   et   al.,  The  proteome  and   transcriptome  analysis   of  Bacillus   subtilis   in  

response  to  salicylic  acid.  Proteomics,  2007.  7(5):  p.  698-­‐710.    42.  Kitko,   R.D.,   et   al.,   Cytoplasmic   acidification   and   the   benzoate   transcriptome   in  

Bacillus  subtilis.  PLoS  One,  2009.  4(12):  p.  e8255.    43.  Riordan,   J.T.,   et   al.,  Response   of   Staphylococcus   aureus   to   salicylate   challenge.   J  

Bacteriol,  2007.  189(1):  p.  220-­‐7.    44.  Chambers,  H.F.  and  F.R.  Deleo,  Waves  of  resistance:  Staphylococcus  aureus  in  the  

antibiotic  era.  Nat  Rev  Microbiol,  2009.  7(9):  p.  629-­‐41.    45.  Kuroda,  M.,  et  al.,  Whole  genome  sequencing  of  meticillin-­‐resistant  Staphylococcus  

aureus.  Lancet,  2001.  357(9264):  p.  1225-­‐40.    46.  Yao,   J.,   et   al.,   Use   of   targetrons   to   disrupt   essential   and   nonessential   genes   in  

Staphylococcus   aureus   reveals   temperature   sensitivity   of   Ll.LtrB   group   II   intron  splicing.  RNA,  2006.  12(7):  p.  1271-­‐81.  

 47.  David,   M.Z.   and   R.S.   Daum,   Community-­‐associated   methicillin-­‐resistant  

Staphylococcus   aureus:   epidemiology   and   clinical   consequences   of   an   emerging  epidemic.  Clin  Microbiol  Rev,  2010.  23(3):  p.  616-­‐87.  

 48.  Kazakova,  S.V.,  et  al.,  A  clone  of  methicillin-­‐resistant  Staphylococcus  aureus  among  

professional  football  players.  N  Engl  J  Med,  2005.  352(5):  p.  468-­‐75.  49.  Lowy,  F.D.,  Staphylococcus  aureus  infections.  N  Engl  J  Med,  1998.  339(8):  p.  520-­‐

32.  50.  Miller,   L.G.   and   B.A.   Diep,   Clinical   practice:   colonization,   fomites,   and   virulence:  

rethinking   the   pathogenesis   of   community-­‐associated   methicillin-­‐resistant  Staphylococcus  aureus  infection.  Clin  Infect  Dis,  2008.  46(5):  p.  752-­‐60.  

 

   66  

51.  Muto,   C.A.,   et   al.,   SHEA   guideline   for   preventing   nosocomial   transmission   of  multidrug-­‐resistant   strains   of   Staphylococcus   aureus   and   enterococcus.   Infect  Control  Hosp  Epidemiol,  2003.  24(5):  p.  362-­‐86.  

 52.  Gill,  S.R.,  et  al.,  Insights  on  evolution  of  virulence  and  resistance  from  the  complete  

genome  analysis  of  an  early  methicillin-­‐resistant  Staphylococcus  aureus  strain  and  a   biofilm-­‐producing   methicillin-­‐resistant   Staphylococcus   epidermidis   strain.   J  Bacteriol,  2005.  187(7):  p.  2426-­‐38.  

 53.  Ito,   T.,   et   al.,   Insights   on   antibiotic   resistance   of   Staphylococcus   aureus   from   its  

whole  genome:  genomic  island  SCC.  Drug  Resist  Updat,  2003.  6(1):  p.  41-­‐52.    54.  Kirby,   W.M.,   Extraction   of   a   Highly   Potent   Penicillin   Inactivator   from   Penicillin  

Resistant  Staphylococci.  Science,  1944.  99(2579):  p.  452-­‐3.    55.  Barber,   M.   and   M.   Rozwadowska-­‐Dowzenko,   Infection   by   penicillin-­‐resistant  

staphylococci.  Lancet,  1948.  2(6530):  p.  641-­‐4.    56.  Klevens,   R.M.,   et   al.,   Invasive   methicillin-­‐resistant   Staphylococcus   aureus  

infections  in  the  United  States.  JAMA,  2007.  298(15):  p.  1763-­‐71.    57.  Hiramatsu,  K.,  Vancomycin  resistance  in  staphylococci.  Drug  Resist  Updat,  1998.  

1(2):  p.  135-­‐50.    58.  Hiramatsu,   K.,   Vancomycin-­‐resistant   Staphylococcus   aureus:   a   new   model   of  

antibiotic  resistance.  Lancet  Infect  Dis,  2001.  1(3):  p.  147-­‐55.    59.  Weigel,  L.M.,  et  al.,  Genetic  analysis  of  a  high-­‐level  vancomycin-­‐resistant  isolate  of  

Staphylococcus  aureus.  Science,  2003.  302(5650):  p.  1569-­‐71.    60.  Ohlsen,  K.,  Novel  antibiotics  for  the  treatment  of  Staphylococcus  aureus  :  Current  

status  of  anti-­‐S.  aureus  antibiotics.  Medscape,  2009.  2(6):  p.  661-­‐672.    61.  Alekshun,   M.N.   and   S.B.   Levy,   Regulation   of   chromosomally   mediated   multiple  

antibiotic   resistance:   the   mar   regulon.   Antimicrob   Agents   Chemother,   1997.  41(10):  p.  2067-­‐75.  

 62.  Martin,   R.G.   and   J.L.   Rosner,   Binding   of   purified   multiple   antibiotic-­‐resistance  

repressor  protein   (MarR)   to  mar  operator   sequences.   Proc   Natl   Acad   Sci   U   S   A,  1995.  92(12):  p.  5456-­‐60.  

 63.  Kaatz,  G.W.,  V.V.  Moudgal,   and  S.M.   Seo,   Identification  and  characterization  of  a  

novel   efflux-­‐related   multidrug   resistance   phenotype   in   Staphylococcus   aureus.   J  Antimicrob  Chemother,  2002.  50(6):  p.  833-­‐8.  

   67  

64.  Kaatz,  G.W.,  S.M.  Seo,  and  C.A.  Ruble,  Efflux-­‐mediated  fluoroquinolone  resistance  in  Staphylococcus  aureus.  Antimicrob  Agents  Chemother,  1993.  37(5):  p.  1086-­‐94.  

 65.  Narui,  K.,  et  al.,  Cloning  and  characterization  of  a  novel  chromosomal  drug  efflux  

gene  in  Staphylococcus  aureus.  Biol  Pharm  Bull,  2002.  25(12):  p.  1533-­‐6.    66.  Truong-­‐Bolduc,  Q.C.,  et  al.,  MgrA  is  a  multiple  regulator  of  two  new  efflux  pumps  

in  Staphylococcus  aureus.  J  Bacteriol,  2005.  187(7):  p.  2395-­‐405.    67.  Bryan,   L.E.   and   S.   Kwan,   Roles   of   ribosomal   binding,   membrane   potential,   and  

electron  transport  in  bacterial  uptake  of  streptomycin  and  gentamicin.  Antimicrob  Agents  Chemother,  1983.  23(6):  p.  835-­‐45.  

 68.  Damper,   P.D.   and   W.   Epstein,   Role   of   the   membrane   potential   in   bacterial  

resistance   to   aminoglycoside   antibiotics.   Antimicrob   Agents   Chemother,   1981.  20(6):  p.  803-­‐8.  

 69.  Price,  C.T.  and  J.E.  Gustafson,  Increases  in  the  mutation  frequency  at  which  fusidic  

acid-­‐resistant  Staphylococcus  aureus  arise  with  salicylate.   J  Med  Microbiol,  2001.  50(1):  p.  104-­‐6.  

 70.  Price,  C.T.,  G.W.  Kaatz,  and  J.E.  Gustafson,  The  multidrug  efflux  pump  NorA  is  not  

required  for  salicylate-­‐induced  reduction  in  drug  accumulation  by  Staphylococcus  aureus.  Int  J  Antimicrob  Agents,  2002.  20(3):  p.  206-­‐13.  

 71.  Riordan,   J.T.,   J.O.   O'Leary,   and   J.E.   Gustafson,  Contributions   of   sigB  and   sarA   to  

distinct  multiple   antimicrobial   resistance  mechanisms   of   Staphylococcus   aureus.  Int  J  Antimicrob  Agents,  2006.  28(1):  p.  54-­‐61.  

 72.  Kaatz,  G.W.,  C.E.  DeMarco,  and  S.M.  Seo,  MepR,  a  repressor  of  the  Staphylococcus  

aureus   MATE   family   multidrug   efflux   pump   MepA,   is   a   substrate-­‐responsive  regulatory  protein.  Antimicrob  Agents  Chemother,  2006.  50(4):  p.  1276-­‐81.  

 73.  Kaatz,  G.W.,  R.V.  Thyagarajan,  and  S.M.  Seo,  Effect  of  promoter  region  mutations  

and   mgrA   overexpression   on   transcription   of   norA,   which   encodes   a  Staphylococcus   aureus   multidrug   efflux   transporter.   Antimicrob   Agents  Chemother,  2005.  49(1):  p.  161-­‐9.  

 74.  O'Leary,   J.O.,   et   al.,   Effects   of   sarA   inactivation   on   the   intrinsic   multidrug  

resistance   mechanism   of   Staphylococcus   aureus.   FEMS   Microbiol   Lett,   2004.  237(2):  p.  297-­‐302.  

 

   68  

75.  Truong-­‐Bolduc,   Q.C.,   J.   Strahilevitz,   and   D.C.   Hooper,   NorC,   a   new   efflux   pump  regulated   by   MgrA   of   Staphylococcus   aureus.   Antimicrob   Agents   Chemother,  2006.  50(3):  p.  1104-­‐7.  

 76.  Truong-­‐Bolduc,  Q.C.,  X.  Zhang,  and  D.C.  Hooper,  Characterization  of  NorR  protein,  

a   multifunctional   regulator   of   norA   expression   in   Staphylococcus   aureus.   J  Bacteriol,  2003.  185(10):  p.  3127-­‐38.  

 77.  Senn,   M.M.,   et   al.,  Molecular   analysis   and   organization   of   the   sigmaB  operon   in  

Staphylococcus  aureus.  J  Bacteriol,  2005.  187(23):  p.  8006-­‐19.    78.  Chan,   P.F.,   et   al.,   The   Staphylococcus   aureus   alternative   sigma   factor   sigmaB  

controls   the   environmental   stress   response   but   not   starvation   survival   or  pathogenicity  in  a  mouse  abscess  model.  J  Bacteriol,  1998.  180(23):  p.  6082-­‐9.  

 79.  Bischoff,  M.,  et  al.,  Microarray-­‐based  analysis  of  the  Staphylococcus  aureus  sigmaB  

regulon.  J  Bacteriol,  2004.  186(13):  p.  4085-­‐99.    80.  Bischoff,  M.,  J.M.  Entenza,  and  P.  Giachino,  Influence  of  a  functional  sigB  operon  on  

the   global   regulators   sar   and   agr   in   Staphylococcus   aureus.   J   Bacteriol,   2001.  183(17):  p.  5171-­‐9.  

 81.  Cheung,   A.L.   and   G.   Zhang,   Global   regulation   of   virulence   determinants   in  

Staphylococcus  aureus  by  the  SarA  protein  family.  Front  Biosci,  2002.  7:  p.  d1825-­‐42.  

 82.  You,   K.,   Salicylate   and  mitochondrial   injury   in   Reye's   syndrome.   Science,   1983.  

221(4606):  p.  163-­‐5.    83.  Gittelman,  D.K.,  Chronic  salicylate  intoxication.  South  Med  J,  1993.  86(6):  p.  683-­‐

5.    84.  Beynen,   A.C.,   et   al.,   Inhibition   of   hepatic   lipogenesis   by   salicylate.   Toxicology,  

1982.  24(1):  p.  33-­‐43.    85.  Oldham,   J.W.,   R.F.   Preston,   and   J.D.   Paulson,   Mutagenicity   testing   of   selected  

analgesics  in  Ames  Salmonella  strains.  J  Appl  Toxicol,  1986.  6(4):  p.  237-­‐43.    86.  Chopra,  I.,  Mechanisms  of  resistance  to  fusidic  acid  in  Staphylococcus  aureus.  J  Gen  

Microbiol,  1976.  96(2):  p.  229-­‐38.    87.  Ferrero,   L.,   et   al.,   Cloning   and   primary   structure   of   Staphylococcus   aureus  DNA  

topoisomerase   IV:   a   primary   target   of   fluoroquinolones.   Mol   Microbiol,   1994.  13(4):  p.  641-­‐53.  

   69  

88.  Cirz,  R.T.,  et  al.,  Complete  and  SOS-­‐mediated  response  of  Staphylococcus  aureus  to  the  antibiotic  ciprofloxacin.  J  Bacteriol,  2007.  189(2):  p.  531-­‐9.  

 89.  Kohanski,   M.A.,   et   al.,   A   common   mechanism   of   cellular   death   induced   by  

bactericidal  antibiotics.  Cell,  2007.  130(5):  p.  797-­‐810.    90.  Udo,   E.E.,   et   al.,   Molecular   characterization   of   epidemic   ciprofloxacin-­‐   and  

methicillin-­‐resistant   Staphylococcus   aureus   strains   colonizing   patients   in   an  intensive  care  unit.  J  Clin  Microbiol,  1996.  34(12):  p.  3242-­‐4.  

 91.  Shalit,   I.,   et   al.,   Widespread   quinolone   resistance   among   methicillin-­‐resistant  

Staphylococcus   aureus   isolates   in   a   general   hospital.   Antimicrob   Agents  Chemother,  1989.  33(4):  p.  593-­‐4.  

 92.  Paez,   P.L.,   M.C.   Becerra,   and   I.   Albesa,   Effect   of   the   association   of   reduced  

glutathione   and   ciprofloxacin   on   the   antimicrobial   activity   in   Staphylococcus  aureus.  FEMS  Microbiol  Lett,  2010.  303(1):  p.  101-­‐5.  

 93.  Gillespie,   S.H.,   et   al.,   Effect   of   subinhibitory   concentrations   of   ciprofloxacin   on  

Mycobacterium  fortuitum  mutation  rates.   J  Antimicrob  Chemother,  2005.  56(2):  p.  344-­‐8.  

 94.  Mesak,   L.R.,   V.   Miao,   and   J.   Davies,   Effects   of   subinhibitory   concentrations   of  

antibiotics   on   SOS   and   DNA   repair   gene   expression   in   Staphylococcus   aureus.  Antimicrob  Agents  Chemother,  2008.  52(9):  p.  3394-­‐7.  

 95.  Aubry-­‐Damon,  H.,  C.J.  Soussy,  and  P.  Courvalin,  Characterization  of  mutations  in  

the   rpoB   gene   that   confer   rifampin   resistance   in   Staphylococcus   aureus.  Antimicrob  Agents  Chemother,  1998.  42(10):  p.  2590-­‐4.  

 96.  Berlanga,   M.   and   M.   Vinas,   Salicylate   induction   of   phenotypic   resistance   to  

quinolones  in  Serratia  marcescens.  J  Antimicrob  Chemother,  2000.  46(2):  p.  279-­‐82.  

 97.  Huang,  J.,  et  al.,  Novel  chromosomally  encoded  multidrug  efflux  transporter  MdeA  

in  Staphylococcus  aureus.  Antimicrob  Agents  Chemother,  2004.  48(3):  p.  909-­‐17.    98.  Wu,   S.,   H.   de   Lencastre,   and   A.   Tomasz,   Sigma-­‐B,   a   putative   operon   encoding  

alternate   sigma   factor   of   Staphylococcus   aureus   RNA   polymerase:   molecular  cloning  and  DNA  sequencing.  J  Bacteriol,  1996.  178(20):  p.  6036-­‐42.  

 99.  Singh,  V.K.,  et  al.,  Impact  of  sigB  mutation  on  Staphylococcus  aureus  oxacillin  and  

vancomycin   resistance   varies   with   parental   background   and   method   of  assessment.  Int  J  Antimicrob  Agents,  2003.  21(3):  p.  256-­‐61.  

   70  

100.   Piriz   Duran,   S.,   F.H.   Kayser,   and   B.   Berger-­‐Bachi,   Impact   of   sar   and   agr   on  methicillin   resistance   in   Staphylococcus   aureus.   FEMS   Microbiol   Lett,   1996.  141(2-­‐3):  p.  255-­‐60.  

 101.   Ashby,  M.J.,  et  al.,  Effect  of  antibiotics  on  non-­‐growing  planktonic  cells  and  biofilms  

of  Escherichia  coli.  J  Antimicrob  Chemother,  1994.  33(3):  p.  443-­‐52.  102.  MSDS,  2012.    103.   Horsburgh,   M.J.,   et   al.,   sigmaB  modulates   virulence   determinant   expression   and  

stress   resistance:   characterization   of   a   functional   rsbU   strain   derived   from  Staphylococcus  aureus  8325-­‐4.  J  Bacteriol,  2002.  184(19):  p.  5457-­‐67.  

 104.   Duthie,   E.S.   and   L.L.   Lorenz,   Staphylococcal   coagulase;   mode   of   action   and  

antigenicity.  J  Gen  Microbiol,  1952.  6(1-­‐2):  p.  95-­‐107.    105.   Luong,  T.T.,  et  al.,  Transcription  Profiling  of  the  mgrA  Regulon  in  Staphylococcus  

aureus.  J  Bacteriol,  2006.  188(5):  p.  1899-­‐910.    106.   Nygaard,   T.K.,   et   al.,   SaeR   binds   a   consensus   sequence   within   virulence   gene  

promoters  to  advance  USA300  pathogenesis.  J  Infect  Dis,  2010.  201(2):  p.  241-­‐54.    107.   Bachmann,   B.J.,   Derivations   and   genotypes   of   some   mutant   derivatives   of  

Escherichia  coli  K12.  American  Society  of  Microbiology,  1987.  2:  p.  1190-­‐1219.  108.   Foster,   P.L.,   Methods   for   determining   spontaneous   mutation   rates.   Methods  

Enzymol,  2006.  409:  p.  195-­‐213.    109.  Matthew  A.  Wikler,  M.,  MBA,   FIDSA   Franklin   R.   Cockerill,   III,  MD,   Karen  Bush,  

PhD  and  P.  Michael  N.  Dudley,  FIDSA  George  M.  Eliopoulos,  MD,  Dwight  J.  Hardy,  PhD  David  W.  Hecht,  MD,   Janet  F.  Hindler,  MCLS,  MT(ASCP)   Jean  B.  Patel,  PhD,  D(ABMM),   Mair   Powell,   MD,   FRCP,   FRCPath   John   D.   Turnidge,   MD,   Melvin   P.  Weinstein,  MD  Barbara  L.   Zimmer,   PhD,  Mary   Jane  Ferraro,   PhD,  MPH   Jana  M.  Swenson,   MMSc,   Methods   for   Dilution   Antimicrobial   Susceptibility   Tests   for  Bacteria   That   Grow   Aerobically;   Approved   Standard-­‐Eight   Edition.   Clinical   and  Laboratory  Standaards  Institute,  2009.  29(2):  p.  10.  

 110.   Edward   A.   Adelberg,   M.M.,   Grace   Chein   Ching   Chen,   Optimal   conditions   for  

mutagenesis   by   N-­‐methyl-­‐N'-­‐nitro-­‐N-­‐nitrosoguanidine   in   Escherichia   coli   K12.  Elsevier,  1965.  18(5-­‐6):  p.  788-­‐795.  

 111.   Schmitz,  F.J.,  et  al.,  Characterization  of  grlA,  grlB,  gyrA,  and  gyrB  mutations  in  116  

unrelated   isolates   of   Staphylococcus   aureus   and   effects   of   mutations   on  ciprofloxacin  MIC.  Antimicrob  Agents  Chemother,  1998.  42(5):  p.  1249-­‐52.  

 

   71  

112.   Oonishi,   Y.,   J.   Mitsuyama,   and   K.   Yamaguchi,   Effect   of   GrlA   mutation   on   the  development   of   quinolone   resistance   in   Staphylococcus   aureus   in   an   in   vitro  pharmacokinetic  model.  J  Antimicrob  Chemother,  2007.  60(5):  p.  1030-­‐7.  

 113.  Martinez,   J.A.,   et   al.,   Analysis   of   grlA   mutations   in   clinical   isolates   of  

Staphylococcus   aureus   with   different   levels   of   quinolone   resistance.   Antimicrob  Agents  Chemother,  1998.  42(5):  p.  1306-­‐7.  

 114.   Ferrero,   L.,   B.   Cameron,   and   J.   Crouzet,  Analysis   of   gyrA   and   grlA  mutations   in  

stepwise-­‐selected   ciprofloxacin-­‐resistant   mutants   of   Staphylococcus   aureus.  Antimicrob  Agents  Chemother,  1995.  39(7):  p.  1554-­‐8.  

 115.   Hernandez,  A.,  M.B.  Sanchez,  and  J.L.  Martinez,  Quinolone  resistance:  much  more  

than  predicted.  Front  Microbiol,  2011.  2:  p.  22.    116.   Tanaka,   T.   and   B.  Weisblum,  Mutant  of   Staphylococcus  aureus  with   lincomycin-­‐  

and   carbomycin-­‐inducible   resistance   to   erythromycin.   Antimicrob   Agents  Chemother,  1974.  5(5):  p.  538-­‐40.  

 117.   Prunier,   A.L.,   et   al.,   Clinical   isolates   of   Staphylococcus   aureus   with   ribosomal  

mutations   conferring   resistance   to   macrolides.   Antimicrob   Agents   Chemother,  2002.  46(9):  p.  3054-­‐6.  

 118.   Pane-­‐Farre,   J.,   et   al.,   The   sigmaB   regulon   in   Staphylococcus   aureus   and   its  

regulation.  Int  J  Med  Microbiol,  2006.  296(4-­‐5):  p.  237-­‐58.    119.   Catherine   M.   Oliphant,   G.M.G.,   Quinolones:   A   Comprehensive   Review.   American  

Family  Physician  2002.  65(3):  p.  455-­‐465.    120.   Chin,   N.X.   and   H.C.   Neu,   Ciprofloxacin,   a   quinolone   carboxylic   acid   compound  

active   against   aerobic   and   anaerobic   bacteria.   Antimicrob   Agents   Chemother,  1984.  25(3):  p.  319-­‐26.  

 121.   Isaacson,   S.H.,   J.   Carr,   and   A.J.   Rowan,   Ciprofloxacin-­‐induced   complex   partial  

status   epilepticus   manifesting   as   an   acute   confusional   state.   Neurology,   1993.  43(8):  p.  1619-­‐21.  

 122.   Schaefler,   S.,   Methicillin-­‐resistant   strains   of   Staphylococcus   aureus   resistant   to  

quinolones.  J  Clin  Microbiol,  1989.  27(2):  p.  335-­‐6.    123.  Warren,   A.,   et   al.,  Multi-­‐drug   resistant   escherichia   coli   with   extended-­‐spectrum  

beta-­‐lactamase   activity   and   fluoroquinolone   resistance   isolated   from   clinical  infections  in  dogs.  Aust  Vet  J,  2001.  79(9):  p.  621-­‐3.  

   72  

124.   Pedraza-­‐Reyes,   M.   and   R.E.   Yasbin,   Contribution   of   the   mismatch   DNA   repair  system  to  the  generation  of  stationary-­‐phase-­‐induced  mutants  of  Bacillus  subtilis.  J  Bacteriol,  2004.  186(19):  p.  6485-­‐91.  

 125.   Harris,   R.S.,   et   al.,   Mismatch   repair   protein   MutL   becomes   limiting   during  

stationary-­‐phase  mutation.  Genes  Dev,  1997.  11(18):  p.  2426-­‐37.    126.   Hobert,  O.,  The  impact  of  whole  genome  sequencing  on  model  system  genetics:  get  

ready  for  the  ride.  Genetics,  2010.  184(2):  p.  317-­‐9.    127.   Lerner,   C.G.,   et   al.,  Novel  approach   to  mapping  of   resistance  mutations   in  whole  

genomes   by   using   restriction   enzyme   modulation   of   transformation   efficiency.  Antimicrob  Agents  Chemother,  2005.  49(7):  p.  2767-­‐77.  

 128.   Vermulst,  M.,  J.H.  Bielas,  and  L.A.  Loeb,  Quantification  of  random  mutations  in  the  

mitochondrial  genome.  Methods,  2008.  46(4):  p.  263-­‐8.    129.   Howden,   B.P.,   et   al.,   Genomic   analysis   reveals   a   point   mutation   in   the   two-­‐

component  sensor  gene  graS  that   leads  to   intermediate  vancomycin  resistance  in  clinical  Staphylococcus  aureus.   Antimicrob  Agents   Chemother,   2008.  52(10):   p.  3755-­‐62.  

 130.   Howden,  B.P.,   et   al.,  Different  bacterial  gene  expression  patterns  and  attenuated  

host  immune  responses  are  associated  with  the  evolution  of  low-­‐level  vancomycin  resistance   during   persistent   methicillin-­‐resistant   Staphylococcus   aureus  bacteraemia.  BMC  Microbiol,  2008.  8:  p.  39.  

 131.   Cui,   L.,   et   al.,   Contribution   of   vraSR   and   graSR   point   mutations   to   vancomycin  

resistance  in  vancomycin-­‐intermediate  Staphylococcus  aureus.  Antimicrob  Agents  Chemother,  2009.  53(3):  p.  1231-­‐4.  

 132.   Neoh,  H.M.,   et   al.,  Mutated  response  regulator  graR  is  responsible  for  phenotypic  

conversion   of   Staphylococcus   aureus   from   heterogeneous   vancomycin-­‐intermediate   resistance   to   vancomycin-­‐intermediate   resistance.   Antimicrob  Agents  Chemother,  2008.  52(1):  p.  45-­‐53.  

 133.  Mitchell,   G.,   et   al.,  A  role   for   sigma   factor  B   in   the   emergence  of   Staphylococcus  

aureus   small-­‐colony   variants   and   elevated   biofilm   production   resulting   from   an  exposure  to  aminoglycosides.  Microb  Pathog,  2010.  48(1):  p.  18-­‐27.  

 134.   Bischoff,  M.  and  B.  Berger-­‐Bachi,  Teicoplanin  stress-­‐selected  mutations  increasing  

sigma(B)  activity  in  Staphylococcus  aureus.  Antimicrob  Agents  Chemother,  2001.  45(6):  p.  1714-­‐20.  

   73  

135.   Berger-­‐Bachi,   B.,   et   al.,  Mapping   and   characterization   of  multiple   chromosomal  factors   involved   in   methicillin   resistance   in   Staphylococcus   aureus.   Antimicrob  Agents  Chemother,  1992.  36(7):  p.  1367-­‐73.  

 136.   Giachino,  P.,  S.  Engelmann,  and  M.  Bischoff,  Sigma(B)  activity  depends  on  RsbU  in  

Staphylococcus  aureus.  J  Bacteriol,  2001.  183(6):  p.  1843-­‐52.    137.   Palma,  M.,   et   al.,  Salicylic  acid  activates  sigma   factor  B  by  rsbU-­‐dependent  and   -­‐

independent  mechanisms.  J  Bacteriol,  2006.  188(16):  p.  5896-­‐903.    138.   Battaglia,   V.,   M.   Salvi,   and   A.   Toninello,   Oxidative   stress   is   responsible   for  

mitochondrial   permeability   transition   induction   by   salicylate   in   liver  mitochondria.  J  Biol  Chem,  2005.  280(40):  p.  33864-­‐72.  

 139.  Martinez,   J.L.   and   F.   Rojo,   Metabolic   regulation   of   antibiotic   resistance.   FEMS  

Microbiol  Rev,  2011.  35(5):  p.  768-­‐89.    140.   Levin,  B.R.  and  D.E.  Rozen,  Non-­‐inherited  antibiotic  resistance.  Nat  Rev  Microbiol,  

2006.  4(7):  p.  556-­‐62.    141.  Martinez,  J.L.,  The  role  of  natural  environments  in  the  evolution  of  resistance  traits  

in  pathogenic  bacteria.  Proc  Biol  Sci,  2009.  276(1667):  p.  2521-­‐30.    142.  McNamara,   P.J.   and   R.A.   Proctor,   Staphylococcus   aureus   small   colony   variants,  

electron  transport  and  persistent  infections.  Int  J  Antimicrob  Agents,  2000.  14(2):  p.  117-­‐22.  

 143.  Meyer,   H.,   M.   Liebeke,   and   M.   Lalk,   A   protocol   for   the   investigation   of   the  

intracellular  Staphylococcus  aureus  metabolome.  Anal  Biochem,  2010.  401(2):  p.  250-­‐9.  

 144.   Copley,  S.D.,  et  al.,  Detection  and  possible  origins  of  aminomalonic  acid  in  protein  

hydrolysates.  Anal  Biochem,  1992.  201(1):  p.  152-­‐7.    145.   Fuchs,   S.,   et   al.,  Anaerobic  gene  expression  in  Staphylococcus  aureus.   J  Bacteriol,  

2007.  189(11):  p.  4275-­‐89.    146.   Rode,  T.M.,  et  al.,  Responses  of  Staphylococcus  aureus  exposed  to  HCl  and  organic  

acid  stress.  Can  J  Microbiol,  2010.  56(9):  p.  777-­‐92.    147.   Porter,  S.N.,  G.S.  Howarth,  and  R.N.  Butler,  Non-­‐steroidal  anti-­‐inflammatory  drugs  

and  apoptosis  in  the  gastrointestinal  tract:  potential  role  of  the  pentose  phosphate  pathways.  Eur  J  Pharmacol,  2000.  397(1):  p.  1-­‐9.  

   74  

148.  Michaels,   M.L.   and   J.H.   Miller,   The   GO   system   protects   organisms   from   the  mutagenic   effect   of   the   spontaneous   lesion   8-­‐hydroxyguanine   (7,8-­‐dihydro-­‐8-­‐oxoguanine).  J  Bacteriol,  1992.  174(20):  p.  6321-­‐5.  

 149.   Hsieh,   S.C.,   et   al.,  The  bacterial  metabolite  2,3-­‐butanediol  ameliorates  endotoxin-­‐

induced  acute  lung  injury  in  rats.  Microbes  Infect,  2007.  9(12-­‐13):  p.  1402-­‐9.    150.  Wright,  G.D.,  On  the  road  to  bacterial  cell  death.  Cell,  2007.  130(5):  p.  781-­‐3.    151.   Chatterjee,   I.,   et   al.,   Staphylococcus   aureus   ClpC   ATPase   is   a   late   growth   phase  

effector  of  metabolism  and  persistence.  Proteomics,  2009.  9(5):  p.  1152-­‐76.    152.   Goswitz,  J.J.,  et  al.,  Detection  of  gyrA  gene  mutations  associated  with  ciprofloxacin  

resistance   in  methicillin-­‐resistant   Staphylococcus   aureus:   analysis   by   polymerase  chain   reaction   and   automated   direct   DNA   sequencing.   Antimicrob   Agents  Chemother,  1992.  36(5):  p.  1166-­‐9.  

 153.   Nagel,  M.,   et   al.,   Influence  of  ciprofloxacin  and  vancomycin  on  mutation  rate  and  

transposition  of  IS256  in  Staphylococcus  aureus.  Int  J  Med  Microbiol,  2010.    154.   Frerman,   F.E.   and   D.C.   White,   Membrane   lipid   changes   during   formation   of   a  

functional  electron  transport  system  in  Staphylococcus  aureus.   J   Bacteriol,   1967.  94(6):  p.  1868-­‐74.  

 155.   Nulton-­‐Persson,   A.C.,   L.I.   Szweda,   and   H.A.   Sadek,   Inhibition   of   cardiac  

mitochondrial   respiration   by   salicylic   acid   and   acetylsalicylate.   J   Cardiovasc  Pharmacol,  2004.  44(5):  p.  591-­‐5.  

 156.   Riordan,  J.T.,  et  al.,  Alterations  in  the  transcriptome  and  antibiotic  susceptibility  of  

Staphylococcus   aureus   grown   in   the   presence   of   diclofenac.   Ann   Clin   Microbiol  Antimicrob,  2011.  10:  p.  30.